Abstract
BACKGROUND: Some autistic people “camouflage” by modifying autistic characteristics in social situations. Operationalizations of camouflaging remain limited and inconsistent. Quantitative studies primarily use retrospective self-report questionnaires, constraining ecological precision and conceptual scope. Methodological limitations obscure knowledge about (1) how multi-faceted camouflaging behaviors and experiences, as part of general impression management (IM), differ between autistic and non-autistic people; and (2) how episodic IM vary by social demands and individual traits. METHODS: We developed a context-discrepancy assessment of camouflaging, conceptualized as a form of IM, in autistic and non-autistic adults. Forty-eight adults (23 autistic, 25 non-autistic) recorded video responses to two hypothetical scenarios with discrepant social evaluative pressure (i.e., job interview vs. video call with trusted other). Immediately after filming, participants rated facets of their IM experiences, including felt inauthenticity, behavioral monitoring extent, effort, and anxiety, and identified behavioral aspects that they particularly monitored. We analyzed IM experiential facet ratings using a 2 (diagnosis) by 2 (social demand) mixed-factorial ANOVA and open-ended data using summative content analysis. Pooled elastic-net regression explored associations between self-reported individual traits (including social coping, cognitive skills, expressivity, gender identity, and neurodivergent traits) and context-dependent discrepancies in IM experiential facet ratings. RESULTS: IM felt more effortful for all participants when social demands increased, but autistic adults experienced greater inauthenticity and behavioral monitoring extent when social demands elevated. Increased social demands also disproportionately heightened anxiety for autistic compared to non-autistic adults. Both groups monitored more behavioral aspects under higher social demands, but autistic adults especially increased their monitoring of nonverbal and idiosyncratic behaviors. Context-discrepancies in IM experiential facet ratings showed distinct associations with trait-level camouflaging use versus subjective self-monitoring and executive functioning abilities, as well as with autistic versus ADHD traits. LIMITATIONS: The video-mediated design may not capture the spontaneity of real-life IM. Findings warrant replication with larger, more diverse cohorts, with counter-balanced or randomized designs. Future research is needed for the current findings to generalize to autistic individuals across sexes and genders, communication and intellectual abilities, as well as ethnicities and cultures. CONCLUSIONS: This novel context-discrepancy assessment suggests that increased social-evaluative demands exacerbate the psychological tolls of IM for autistic compared to non-autistic adults. Autistic relative to non-autistic adults may particularly monitor nonverbal behaviors in such circumstances. Individual trait associations further delineate IM facets and implicate cognitive and neurodivergent characteristics as potential sources of heterogeneous IM experiences. These insights shed new light into how autistic and non-autistic adults similarly and differently coordinate and experience social coping in context-dependent ways. SUPPLEMENTARY INFORMATION: The online version contains supplementary material available at 10.1186/s13229-026-00715-2.