Gestalt therapy is a humanistic-experiential and relational psychotherapy that understands psychological experience as emerging within an organism-environment field. Its practice emphasises phenomenological inquiry, dialogical contact, embodiment, experimentation, and attention to the situation in which suffering and change occur ([58]; [93]). This theoretical identity creates both a research opportunity and a research difficulty. The approach proposes clinically meaningful processes that can be studied, but it resists reduction to a fixed collection of techniques. Consequently, empirical work must distinguish between evaluating Gestalt therapy as a coherent model and evaluating an isolated technique, an adapted counselling programme, or a composite intervention that incorporates selected Gestalt elements.

The state of the evidence matters beyond a professional dispute about recognition. Claims of effectiveness influence training, public communication, referral, health-service decisions, and the expectations of clients. A psychotherapy community therefore has an ethical obligation to distinguish clinical plausibility from observed change, observed change from causal effect, and causal effect from comparative or disorder-specific efficacy. This distinction is especially important in small literatures, where a series of favourable studies can create an impression of cumulative confirmation even when samples, interventions, outcomes, and designs are not comparable.

[65] systematic review, published in the Open Journal of Social Sciences, identified 11 empirical studies from the preceding 12 years. It concluded that Gestalt interventions appeared promising, particularly in group settings, while emphasising scarcity and methodological heterogeneity. Since that review, the empirical landscape has changed. A Gestalt Therapy Fidelity Scale has been developed ([21]); larger practice-based cohorts have appeared ([11]; [86]); process and client-experience studies have expanded ([25]; [28]; [71]); and theory-derived measures have been developed for aesthetic relational knowing, therapist responsiveness, interoception, and treatment recognition ([69]; [73]; [79]).

The resulting corpus is substantially larger, but the appropriate conclusion is not simply that evidence has accumulated. The more important question is what kind of evidence has accumulated and what inferences it can sustain. A pre-post group study, a randomised wait-list comparison, a phenomenological interview study, a fidelity-scale validation, and a training survey are all empirical, but they answer different questions. Treating them as interchangeable items in a single body of “evidence for Gestalt therapy” obscures rather than clarifies the state of knowledge.

This article therefore adopts a critical evidence-map approach. It asks: 1) what publicly traceable empirical reports were published between 2019 and July 2026; 2) how the corpus is distributed across outcomes, process, experience, fidelity, measurement, and training; 3) which recurring methodological and epistemic limitations restrict interpretation; and 4) what contemporary research can—and cannot—support. The aim is not to rank Gestalt therapy against other approaches or calculate a pooled treatment effect. It is to calibrate claims to the actual inferential strength of the research.

A critical evidence map was constructed for primary empirical reports assigned to the period 1 January 2019 through July 2026. Evidence mapping was selected because the literature spans heterogeneous questions and designs rather than a single intervention-outcome comparison. The work was informed by PRISMA-ScR, PRISMA-S, and JBI guidance on transparent evidence synthesis ([59]; [67]; [89]), but it is not presented as a completed JBI scoping review.

Searches combined PubMed/MEDLINE and PubMed Central, multidisciplinary academic search systems, the Gestalt Psychotherapy Research Database, DOAJ, publisher and journal archives, DOI records, exact-title and author searches, backward reference checking, forward citation searching, and multilingual keyword searches. Journal-by-journal inspection included Gestalt Review, British Gestalt Journal, Research in Psychotherapy, Frontiers journals, The Humanistic Psychologist, Journal of Humanistic Counseling, and regional journals in which psychotherapy and counselling studies were identified. The corpus was cross-checked against [65] and the later semi-systematic review by [90]. Complete platform-specific strings, interfaces, execution dates, limits, and the projects retained from each verification source are reproduced in Supplementary Data Workbook Sheet S1_Search_Log¹. The final evidence cut-off was 26 July 2026. Verification reruns were conducted on 8 August 2026 using the same cut-off; records published after the cut-off were not added.

Identifiable Gestalt-therapy journals and association archives were inspected in English, Portuguese, Spanish, Italian, German, French, Serbian and Japanese, supplemented by Korean- and Chinese-language journal/catalogue searching. The sweep included, among others, British Gestalt Journal, Gestalt Review, Revista da Abordagem Gestáltica, IGT na Rede, Revista de Gestalt (Instituto Sedes Sapientiae), Revista de Terapia Gestalt (AETG), Quaderni di Gestalt, GTK-Gestalt Therapy Kairos, Gestalt (France), GESTALTTHERAPIE, Geštalt Zbornik, Gestalt Today Malta/EAPTI-GPTIM, the Gestalt Journal of Australia and New Zealand, Japanese Journal of Gestalt Therapy, Japanese Journal of Clinical Gestalt Therapy, and Korean specialist Gestalt sources. Chinese-language intervention searches used local terms for Gestalt therapy and outcome research. Because eligible Gestalt research also appears outside specialist outlets, the audit was supplemented by broader multilingual external-journal searching, including Portuguese-, Spanish-, and Chinese-language search families and exact-title/citation verification in other languages. A journal was recorded as searched even when it yielded no eligible report. Because no authoritative global registry of Gestalt-therapy journals exists, this procedure is described as a dedicated multilingual specialist-journal sweep plus broader multilingual searching rather than a claim that every Gestalt journal or report worldwide was identifiable. Journal-by-journal sources, search approaches, retained records, and exclusions are documented in Supplementary Data Workbook Sheets S1_Search_Log and S1_Journal_Sweep.

The search was iterative rather than derived from a single exportable database query. Core identity concepts included “Gestalt therapy,” “Gestalt psychotherapy,” “Gestalt counseling,” “Gestalt counselling,” “Gestalt-based,” and “Gestalt play therapy.” These were combined in separate searches with outcome and design terms such as “effectiveness,” “efficacy,” “outcome,” “intervention,” “controlled,” “randomised,” “quasi-experimental,” “pre-post,” “cohort,” and “single case”; process and experience terms such as “contact,” “embodiment,” “relationality,” “client experience,” “therapist experience,” “dialogue,” “significant events,” and “case formulation”; and measurement and training terms such as “fidelity,” “adherence,” “competence,” “supervision,” “training,” “interoception,” “responsiveness,” and “aesthetic relational knowing.” Searches were supplemented by local-language variants, exact-title and DOI searches, author searches, journal-volume browsing, backward reference checking, and forward citation tracing. The complete reproducible strings used in the verification reruns are reported in Supplementary Data Workbook Sheet S1_Search_Log.

Several public search interfaces used during the initial iterative identification process did not preserve stable, exportable result sets. Consequently, the original numbers retrieved, automated duplicates, title-level exclusions, and the exclusive discovery source for every report cannot be reconstructed with the precision required for a conventional PRISMA flow diagram. These quantities were not estimated retrospectively. Supplementary Data Workbook Sheet S1_Search_Log distinguishes the original search families from the documented verification reruns. This limitation is why the study is described as a critical evidence map and a publicly traceable minimum corpus rather than as a completed systematic or JBI scoping review.

Eligible reports presented primary quantitative, qualitative, mixed-method, psychometric, implementation, single-case, multiple-case, supervision, or training data explicitly grounded in Gestalt therapy, Gestalt psychotherapy, Gestalt-based counselling, or a specifically Gestalt-derived construct. Reviews, protocols without results, theoretical essays, conference abstracts, nonclinical applications, and clinical illustrations without an explicit research method were excluded from the primary corpus. Chairwork and empty-chair studies were excluded from the core corpus when the surrounding treatment was not identified as Gestalt therapy. One study of cognitive-behavioural treatment integrated with empty-chair work was retained separately as a boundary case because it illustrates why a technique historically associated with Gestalt therapy cannot automatically be treated as evidence for the complete model. This report contributed neither to the 81-report core corpus nor to the 40 core outcome reports.

Reports were classified by primary evidence domain—outcome/implementation, process/experience, or measurement/fidelity/training—and by intervention status: full Gestalt therapy; full/adapted Gestalt therapy; adapted Gestalt therapy; composite intervention; Gestalt-derived measurement or process research; mixed-orientation study with a Gestalt component; or adjacent technique. Multiple reports arising from the same participants or programme were linked as one project to prevent double counting.

The final core map retained 81 reports representing 77 identifiable projects. Forty core reports addressed treatment outcomes or implementation, 23 investigated therapeutic process or lived experience, and 18 concerned fidelity, measurement, training, supervision, or Gestalt-derived competencies. One additional outcome report was retained separately as a boundary case involving an adjacent technique. When the core corpus and boundary case are displayed together for audit purposes, the complete inventory contains 82 reports representing 78 projects.

Unless explicitly stated otherwise, conclusions concerning the empirical support for Gestalt therapy are based on the 81-report core corpus. The boundary case is used only to demonstrate the conceptual and classificatory problem created when an isolated technique is treated as equivalent to the complete treatment model.

The expansion is nevertheless conceptually uneven. Of the 40-core outcome or implementation reports, only 9 were classified as clearly evaluating full Gestalt therapy. Fourteen involved full/adapted Gestalt therapy, 11 evaluated adapted forms, and 6 investigated composite programmes. The separate boundary case concerned an adjacent technique and was not counted as evidence for full or adapted Gestalt therapy. Thus, most core outcome reports cannot be treated as direct replications of a common intervention. The label “Gestalt” covers materially different treatment packages, doses, populations, practitioner qualifications, and settings.

The final multilingual expansion added 25 core reports that were not present in the earlier 56-report audit. Twenty were recovered through specialist Gestalt journals or association archives (4 outcome/implementation, 11 process/experience, and 5 measurement/training reports), and five additional outcome/implementation reports were identified through broader multilingual searches of nonspecialist journals. The latter included a Chinese controlled adjunct intervention ([63]), Mexican child and university-student studies ([64]; [45]), and Brazilian short-term online and prison-counselling reports ([14]; [72]). The specialist/association additions also include Brazilian single-case and process studies ([17]; [8]; [47]; [74]), Spanish outcome and process studies ([51]; [52]; [66]; [70]), Italian, Serbian and Japanese research, and a second linked Japanese report from the six-month group-therapy project, which added quantitative outcome assessment without creating a new independent project ([34]). These additions widen linguistic and geographical coverage but remain dominated by small, local, or context-specific designs rather than large confirmatory trials.

The expanded report-level dataset permits each corpus-level statement to be traced to a bibliographic and content-verification source. It also separates missing reporting from demonstrated absence and makes visible the evidence supporting each intervention-status decision, comparator classification, sample-size value, and direction-of-effect code. The composition and inferential function of the mapped corpus are summarized in Table 1.

Table 1. Composition and inferential function of the mapped corpus.

Outcome reports covered anxiety and mood difficulties, trauma, grief, intimate partner violence, chronic illness, breast and lung cancer, multiple sclerosis, social anxiety, parenting, childhood behavioural problems, hearing impairment, dialysis, marital distress, routine clinical practice, adolescent relational well-being during COVID-19, bereavement, childhood sexual-abuse sequelae, depressive-disorder social functioning, psychological care for displaced children, university-student depression, short-term online care for social isolation, prison psychological-duty implementation, and a six-month Japanese Gestalt group-therapy project. Several reports used active comparators or control groups, including reality therapy, psychodrama, behavioural activation, cognitive-behavioural or acceptance-based interventions, positive psychology, Lego-based programmes, other play therapies, routine care, and no-treatment/control conditions (e.g., [50]; [55]; [63]; [95]). Larger naturalistic cohorts were reported for Gestalt Pastoral Care and routine clinical programmes ([11]; [86]).

Most, but not all, reports described favourable changes. One randomised wait-list trial of Gestalt-based counselling for pregnant women did not establish statistically significant improvement in self-esteem or intimate partner violence ([50]). For the linked Japanese six-month group project, publication identity, participant structure, treatment duration, and assessment design could be verified, but the publicly accessible article record did not provide numerical results; its direction of effect is therefore coded as not assessable rather than inferred ([34]). For reports showing favourable change, the meaning of that change varies sharply by design. Uncontrolled pre-post improvement is compatible with treatment benefit, but it is also compatible with regression to the mean, spontaneous improvement, expectancy, concurrent care, repeated testing, and nonspecific therapeutic attention. Wait-list comparisons provide stronger evidence than uncontrolled change but commonly generate larger effects than credible active comparators ([15]). Small comparative studies can identify promising signals, yet they rarely provide stable disorder-specific effect estimates or robust tests of equivalence.

Process research produced some of the most conceptually specific contributions. Dialogue and phenomenological studies examined moment-to-moment interaction, relationality, bodily awareness, emotional integration, self-regulation, agency, and perceived change ([12]; [25]; [29]; [28]). The expanded multilingual corpus additionally documents dialogical practice among Brazilian Gestalt clinicians, a single-case study of psychological work in multidisciplinary diabetes care, and a qualitative child sexual-abuse case study with an explicit research method ([8]; [17]; [47]), therapists’ experience of online practice in Brazil and the Balkans ([74]; [84]), relational-field and transpersonal-process research ([60]; [87]), therapeutic presence and alliance in Spain ([66]; [70]), and participant experiences in Japanese Gestalt groups ([32]). Studies of therapist practice continued to show substantial variation within the same theoretical orientation and developed empirically grounded accounts of case formulation ([71]; [82]).

These studies matter because they test the conceptual language of the approach and illuminate how therapy is experienced. Their limitation is not that they are qualitative. It is that their findings answer process and meaning questions, not comparative efficacy questions. A coherent evidence synthesis must preserve that distinction. Qualitative depth cannot be converted into an effect estimate, while a symptom score cannot by itself demonstrate contact, embodiment, agency, or relational transformation.

The Gestalt Therapy Fidelity Scale is a pivotal methodological development because it allows observers to distinguish Gestalt from non-Gestalt sessions with acceptable reliability ([21]). A second research line has operationalised aesthetic relational knowing and therapist responsiveness, including comparisons between trainees and experienced clinicians and validation across therapeutic orientations ([79]). The enlarged corpus also includes studies of becoming a Gestalt psychotherapist, competencies and shame in training, compassion fatigue and satisfaction among Gestalt psychotherapists, training-course experience, and a Gestalt-informed psycho-relational assessment in paediatric care ([2]; [10]; [31]; [10]; [68]). Other studies examined embodied supervision, expressive-media training, interoceptive awareness, and the recognition of treatment interventions ([18]; [46]; [69]; [73]).

This infrastructure is arguably the strongest cumulative achievement of the period. It makes future testing more precise. Yet measurement development is not evidence that the measured construct causes client improvement. Theory-derived scales require behavioural validation, convergent and discriminant evidence, temporal testing, and prospective association with client outcomes, dropout, alliance, and unwanted effects before they can be described as mechanisms of change ([35]).

The most fundamental limitation is ontological before it is statistical. The corpus does not contain one stable intervention. “Gestalt therapy” is used for full psychotherapy, brief counselling, play therapy, positive-parenting programmes, pastoral care, mindfulness-Gestalt groups, focusing and arts-based programmes, Gestalt plus systemic pedagogy, and interventions containing empty-chair procedures. Some of these may be clinically coherent and valuable, but their effects cannot automatically be attributed to the same treatment model. This ambiguity produces construct contamination. A positive result for a composite programme may support that programme but not the unique contribution of Gestalt therapy. A successful empty-chair procedure may support chairwork in a particular context but not the full Gestalt model. Conversely, a negative result in an underspecified adaptation should not be treated as a failure of the complete approach. The availability of the Gestalt Therapy Fidelity Scale raises the expected standard: treatment identity should be demonstrated through observed sessions rather than inferred from a label or a list of techniques ([21]).

A recurrent problem is the movement from “participants improved” to “the intervention was effective”. Causal attribution requires a design capable of excluding plausible alternatives. Uncontrolled studies and simple pre-post comparisons cannot isolate treatment-specific effects. Randomisation strengthens inference, but small samples, unclear allocation procedures, unblinded outcome collection, and weak control conditions can still produce unstable or inflated results. Comparative psychotherapy research also requires balanced treatment credibility, competent delivery in each condition, and attention to researcher allegiance ([48]; [97]).

The problem is not solved by treating randomisation as a binary quality marker. A small, randomised wait-list trial may be less informative about relative efficacy than a carefully designed pragmatic study with an active comparator, fidelity checks, repeated outcomes, and transparent attrition. The inferential question must determine the design. Feasibility, acceptability, change, superiority, non-inferiority, mechanism, and implementation are different claims and require different evidence.

Using the prespecified project-level extraction rules, 34 unique quantitative core outcome or implementation projects contributed a verifiable numerical participant total; their median sample size was 43.5. Among 22 projects with two or more concurrent conditions—including active, routine-care, wait-list, placebo, or no-treatment controls—the median sample size was 60. Linked publications from the same project were counted once, and qualitative-only outcome/implementation reports were excluded from these medians. The contributing values and spreadsheet formulas are supplied in Supplementary Data Workbook Sheet S2_Sample_Summary.

Large effects in small local studies should therefore be read as hypotheses requiring replication rather than as settled estimates. The issue is compounded when multiple outcomes are tested, when protocols are adapted for a single population, and when findings are not reproduced by an independent team. Replication is not a rejection of contextual sensitivity; it is the procedure by which the field learns which findings travel across therapists, institutions, and cultures.

Throughout this appraisal, non-reporting was distinguished from demonstrated absence. When a publication did not describe fidelity monitoring, assessor masking, preregistration, clinically significant change, attrition reasons, or adverse-event assessment, the appropriate conclusion was that the feature could not be verified from the report—not that it definitively did not occur. This distinction is particularly important in a dispersed literature characterised by heterogeneous reporting standards. Claims concerning these domains are therefore framed primarily as limitations of the published evidence rather than as categorical statements about the conduct of every study.

Many reports rely on self-report scales administered before and immediately after intervention. Statistical significance establishes that an observed difference is unlikely under a specified null model; it does not establish that clients experienced reliable, clinically meaningful recovery. Psychotherapy outcome research has long distinguished statistical change from reliable and clinically significant change ([26]). The corpus seldom reports recovery, deterioration, reliable change, functional improvement, or the proportion of clients crossing a clinically relevant threshold.

Outcome selection also shapes the story a study can tell. Measures of symptoms, well-being, self-efficacy, empathy, marital adjustment, or personality dimensions are not interchangeable. Multiple measures can enrich interpretation, but they also increase the risk of selective emphasis when outcomes and analyses are not preregistered. Independent or blinded outcome assessment was not consistently described, and the same participant frequently supplies intervention exposure, expectancy, and outcome information, increasing common-method bias.

Therapist affiliation with Gestalt therapy does not demonstrate that a recognisably Gestalt treatment occurred. [71] showed substantial variation in how Gestalt therapists conceptualised and treated similar clinical presentations. Such variation may be a legitimate expression of responsiveness, but it complicates interpretation when session content is not observed. Fidelity and competence should be assessed without turning relational practice into a rigid script.

Therapist effects are another unresolved source of variance. Outcomes may reflect therapist skill, responsiveness, expectations, supervision, or setting as much as the named model. Research on alliance and contextual factors shows that relationship processes are associated with outcomes across orientations ([20]; [91]). Gestalt research needs to determine whether its specific processes add explanatory value beyond these common and therapist-level factors.

Post-treatment assessment was much more consistently described than repeated longer-term follow-up. Supplementary Data Workbook Sheet S3_Appraisal codes follow-up report by report. Follow-up was present in several reports, but its timing, completeness, and analytical treatment varied substantially. Consequently, the durability, delayed emergence, or later reversal of reported changes usually cannot be determined from the available publications.

Attrition was likewise reported inconsistently. Supplementary Data Workbook Sheets S2_Report_Dataset and S3_Appraisal distinguish, where verifiable, the number enrolled or allocated, the number analysed, reported losses, and the handling of missing observations. Many reports did not explain reasons for withdrawal or the analytical consequences of missing data. Withdrawal from psychotherapy may reflect improvement, practical barriers, dissatisfaction, poor fit, treatment burden, deterioration, or an adverse relational experience; reporting only completers therefore limits both effectiveness and safety interpretation.

Explicit assessment and reporting of adverse events, deterioration, and other negative effects were uncommon. The revised dataset distinguishes explicit assessment, reported deterioration, and “not reported or not verifiable.” A small number of reports supplied deterioration or worsening information, including reliable-deterioration categories in routine CORE-OM data and worsening body-image difficulties in a composite trauma programme. This evidence is too sparse and heterogeneous to characterise the safety profile of Gestalt and Gestalt-based interventions. The absence of reported adverse events should not be interpreted as evidence that no adverse events occurred.

Psychotherapy can be associated with unwanted events, treatment-emergent reactions, deterioration, dependency, relational harm, or opportunity costs, even when treatment is competently and benevolently delivered ([42]). A humanistic and relational orientation should be particularly attentive to clients’ accounts of burden, disconfirmation, shame, coercion, rupture, and harm. The absence of reported adverse events should therefore not be interpreted as evidence that no adverse events occurred.

The corpus includes work from Europe, Latin America, North America, the Middle East, Africa, and Asia. This geographical breadth corrects the assumption that credible evidence is produced only in dominant English-language centres. It also demonstrates the social and cultural adaptability of Gestalt-informed practice. However, many studies are closely tied to one institution, one practitioner network, one gender, one age group, or one clinical and cultural setting. Local relevance should not be converted into universal generalisation.

Reporting of diagnosis, comorbidity, medication, socioeconomic position, race or ethnicity, disability, therapist demographics, and service context is inconsistent. These omissions limit judgments about equity and transferability. Cultural adaptation must be described as part of the intervention rather than treated as an invisible background variable.

The corpus contains at least three distinct knowledge systems. Outcome studies ask whether change occurred under particular conditions. Process and phenomenological studies ask how therapy is experienced and organised. Fidelity, training, and measurement studies ask whether theoretical constructs and professional competencies can be operationalised. Each is valuable; none substitutes for the others. A scale validation is not an outcome trial, a client narrative is not a causal comparison, and a statistically significant symptom reduction does not validate a theory of change.

The field’s evidence should therefore be presented as a structured portfolio rather than a single ladder culminating in an undifferentiated statement that “Gestalt therapy works”. The relevant claims are narrower: some Gestalt and Gestalt-based interventions are feasible and associated with favourable change; some controlled studies report context-specific advantages; process research documents theoretically relevant experiences; and measurement research has improved the capacity to study treatment identity and therapist responsiveness.

Much of the literature is produced by clinicians, trainers, or institutions committed to the approach. Such involvement is not inherently disqualifying; expert knowledge is often necessary to design a faithful intervention. Nevertheless, psychotherapy comparisons are vulnerable to allegiance effects, particularly when treatment integrity is not independently assessed and when authors interpret ambiguous results in favour of a preferred model ([48]; [97]). Independent replication and transparent declaration of professional allegiance would strengthen credibility.

The corpus is dispersed across specialist and regional journals, association publications, national repositories, and local-language outlets, many with limited or inconsistent indexing. The final specialist-journal sweep, and broader multilingual external-journal searches materially altered the map, demonstrating that retrieval inequality is not a peripheral concern: research in Portuguese, Spanish, Italian, Serbian, Japanese, and Chinese-language sources can be missed by searches centred on dominant databases. Broader multilingual and journal-level searching therefore improves recall, but it does not remove the need for critical appraisal. Inclusion should broaden the evidence base and the range of contexts represented; it should not lower standards of reporting, verification, or inference.

The corpus changes the empirical portrait of Gestalt therapy. Eighty-one core reports, together with one separately identified boundary case, constitute a materially different empirical landscape from the 11 studies identified in 2019. The field has developed a research ecology involving outcome monitoring, comparative intervention studies, systematic cases, qualitative process analysis, online therapy, supervision, training, fidelity, and psychometric work. Describing Gestalt therapy as wholly unresearched is no longer accurate; describing the literature as well consolidated would be equally misleading.

The expanded corpus, however, does not authorise a broad declaration of established efficacy. The most serious limitations are not merely a shortage of studies but recurrent mismatches between labels, designs, measures, and conclusions. The literature frequently moves across levels of claim: improvement becomes effectiveness; effectiveness becomes model-specific efficacy; one population becomes a general indication; a technique becomes the therapy; a theory-derived measure becomes a mechanism. Critical appraisal requires these transitions to be made explicit and resisted unless the evidence warrants them.

Gestalt theory rightly challenges the assumption that a person can be understood outside a relational and situational field. This does not require rejecting measurement, comparison, or causal reasoning. It requires choosing methods that preserve clinically relevant complexity while answering a defined question. Routine outcome monitoring can examine real-world change; systematic case designs can test temporal patterns within persons; qualitative inquiry can study lived experience and relational process; fidelity measures can establish treatment identity; and pragmatic trials can estimate comparative effects in services. Methodological pluralism is rigorous when each method is matched to the inference it can support.

The alternative—using contextual complexity as a reason to accept vague treatment definitions or uncontrolled conclusions—protects theory from testing rather than studying it. A field perspective can become a research resource: it encourages measurement of therapist, service, culture, relationship, and situation, not only the isolated patient. The strongest future research will not imitate a narrow pharmacological trial model, but neither will it exempt Gestalt claims from causal and evidentiary standards.

The language used by institutes, trainers, practitioners, and authors should reflect the evidence. The contemporary literature supports statements that Gestalt and Gestalt-based interventions have been implemented in diverse settings; that many studies report favourable changes; that some controlled studies identify context-specific advantages; and that theory-specific process and measurement research is developing. It does not support statements that Gestalt therapy is broadly proven across diagnoses, superior to established psychotherapies, or uniformly delivered across practitioners and countries.

This evidence map has limitations that qualify its numerical and interpretive findings. Supplementary Data Workbook Sheets S1_Search_Log and S1_Journal_Sweep supply reproducible search families, documented verification reruns, a dedicated multilingual specialist-journal audit, and broader multilingual external-journal searches. Nevertheless, the yields and screening decisions from the initial iterative searches cannot be reconstructed as if they had been prospectively exported, and no authoritative global registry of Gestalt-therapy journals exists. The international expansion therefore reduces—but cannot eliminate—the possibility that poorly indexed, discontinued, local, inaccessible, or otherwise undiscoverable publications remain unidentified. Public accessibility, catalogue quality, language, and the visibility of professional networks may still influence retrieval. The reported corpus should be understood as a publicly traceable minimum rather than a demonstrably exhaustive worldwide census.

Screening, eligibility decisions, data extraction, and intervention-status classification were conducted by one author and were not independently duplicated. Study-level rationales in Supplementary Data Workbook Sheet S2_Report_Dataset and structured domain coding in Sheet S3_Appraisal improve transparency but do not remove interpretive error. Separate validated risk-of-bias instruments were not applied to every design; the appraisal is therefore structured and design-sensitive but not a formal comparative quality assessment. Some reports were accessible only through abstracts or publisher records. For those reports, unsupported methodological details were not inferred, and unavailable fields were marked as not reported or not verifiable. These limitations require caution regarding exact prevalence estimates, claims of completeness, and comparisons among evidence domains.

Contemporary Gestalt therapy research is substantially larger, more international, and more methodologically diverse than the literature described in 2019. Within the publicly traceable literature mapped here, 81 reports constituted the core corpus, representing 77 identifiable projects. One additional adjacent-technique report was retained as a boundary case; the complete displayed inventory therefore contains 82 reports representing 78 projects. Their distribution across outcome, process, experience, fidelity, measurement, training, and supervision research demonstrates that Gestalt therapy can no longer accurately be described as an approach without empirical investigation.

Numerical expansion, however, should not be confused with evidentiary consolidation. The contemporary literature does not constitute a unified body of replications testing one stable and consistently delivered intervention. It consists of a heterogeneous portfolio that includes full Gestalt psychotherapy, adapted counselling, play therapy, pastoral care, composite programmes, theory-derived measures, qualitative process studies, and interventions incorporating isolated Gestalt-associated techniques. This diversity is intellectually and clinically productive, but it limits the extent to which favourable findings can be accumulated under the general proposition that “Gestalt therapy works”.

The corpus supports several carefully bounded conclusions. Gestalt and Gestalt-informed interventions have been implemented across diverse populations, cultures, and service settings. Many studies report favourable changes, and some controlled comparisons provide preliminary evidence of context-specific benefit. Process research identifies experiences and relational patterns consistent with Gestalt concepts, including embodiment, contact, awareness, responsiveness, self-regulation, and agency. Measurement and fidelity studies have also created an emerging infrastructure through which treatment identity and theory-derived clinical competencies can be examined more precisely.

The corpus does not yet support broad claims that Gestalt therapy has established comparative efficacy across diagnoses, that it is superior or equivalent to established psychotherapies, or that its proposed relational and embodied processes have been demonstrated to mediate clinical change. Such conclusions are restricted by intervention ambiguity, small and selective samples, inconsistent comparison conditions, limited use of fidelity assessment, scarce independent replication, short follow-up, and minimal reporting of deterioration or unwanted effects. Statistical improvement after treatment should therefore not be interpreted automatically as clinically significant, treatment-specific, durable, or generalisable change.

The strongest conclusion is neither that Gestalt therapy has been empirically validated as a homogeneous treatment nor that the existing research is too weak to matter. Rather, the field has reached an intermediate stage of empirical development: it possesses meaningful signals of feasibility, change, process relevance, and conceptual operationalisation, but it has not yet converted those signals into a stable cumulative evidence base. The appropriate response is not to impose a single methodological hierarchy, but to maintain a disciplined correspondence between research design and the claim being made.

This conclusion also has an ethical and institutional dimension. Training institutes, professional associations, researchers, and practitioners should communicate the evidence without either minimising its development or exaggerating its implications. A mature research culture is demonstrated not by the number of favourable findings it can assemble, but by its capacity to distinguish observation from causation, theory from mechanism, a technique from a complete model, and local relevance from general efficacy. On that criterion, the principal task facing contemporary Gestalt therapy research is not simply to generate more positive studies, but to make treatment identity, uncertainty, negative findings, contextual conditions, and the limits of inference visible. The complete report-level corpus matrix is presented in Appendix A.

The author acknowledges the constructive editorial feedback received on an earlier version of this manuscript.

The complete revision dataset is supplied in the accompanying Final International Supplementary Data Workbook: Gestalt Therapy Evidence Map, 2019-2026. It contains the reproducible search log, multilingual specialist-journal sweep, broader multilingual external-journal search record, report-level verification and extraction dataset, project-level sample-size calculations, structured appraisal, excluded/ borderline candidates, and coding dictionary.

S.H.B.T. conceived and designed the review, developed the eligibility and classification framework, identified and verified the literature, extracted and analysed the data, prepared the tables and appendix, and drafted and revised the manuscript.

OpenAI’s ChatGPT was used for language editing and structured data checking. Consensus was used as a bibliographic discovery aid.

The condensed matrix below reproduces the 81 publicly traceable reports included in the core evidence map and one separately identified boundary-case report. The boundary case is displayed to make the inclusion boundary visible but is excluded from core totals and from claims concerning the empirical support for Gestalt therapy. “Status” describes the relation of the study to the complete Gestalt model; it is not a quality rating. Reports sharing the same project identifier were treated as linked publications and participant counts were not added across those reports. The complete source-level extraction, access level, classification rationale, outcome fields, and appraisal are supplied in the Supplementary Data Workbook.

Table A. Core empirical reports and boundary case, 2019-2026.

Continued

Continued

Continued

Continued

¹Supplementary Data Workbook is published on Mendeley Data. https://doi.org/10.17632/xvzm7cs7wh.1