We aimed to investigate prospective psychotherapists' bias in assessing their own facilitative interpersonal skills (FIS) and predictors of high self-assessments. In this cross-sectional observational study, we examined 132 psychology students and trainee psychotherapists. Therapists' demographic variables and self-concepts were assessed through self-report questionnaires, and their therapeutic skills were assessed with the German version of the "FIS" task. A truth-and-bias model approach was applied in three different hierarchical linear models to test self-assessment bias of FIS, and to identify factors associated with overly positive self-assessments of therapeutic skills. Significantly higher self-assessments of skills were found in comparison to observer ratings for overall FIS and for the FIS dimensions hope, emotional expression, warmth, acceptance, and understanding, empathy, alliance bond capacity, and rupture-repair responsiveness. Despite this discrepancy, there was a significant congruence between self- and observer ratings. A greater self-assessed ego-strength and attractiveness were associated with self-assessments of skills that were higher than the observer-rated average. Age, gender, experience, and other self-concept variables showed no significant association with self-assessments. This present study demonstrates self-assessment bias in prospective therapists. However, when participants rated their own FIS to be high, observers also tended to rate participants' skills higher. Especially therapists who generally feel more confident and attractive tend to estimate themselves positively. Supervisors and trainees should be aware of the ubiquity of self-assessment bias. Future research is necessary to test the effect of self-assessment bias on patient outcomes and process variables. (PsycInfo Database Record (c) 2023 APA, all rights reserved).