Objective: The present research examined whether concurrent expert testimony, or hot tubbing, is able to reduce adversarial allegiance compared to traditional adversarial expert testimony. Hypotheses: We expected concurrent experts would display less adversarial allegiance over the course of a mock criminal responsibility evaluation, with case opinions converging between prosecution and defense witnesses. We expected adversarial experts would display escalating adversarial allegiance with increasingly divergent case opinions and that court-appointed experts’ opinions would remain stable over the evaluation. We also expected concurrent expert witnesses would produce more balanced expert reports and testimony than would adversarial experts. Method: Clinicians and advanced clinical doctoral students (N = 103) completed criminal responsibility evaluation training before conducting a mock criminal responsibility evaluation for the prosecution, defense, or the court. Half of the partisan experts followed traditional adversarial procedures and half followed a concurrent testimony process. Participants provided case perceptions at three time points: after initial evidence review, after completing expert report(s), and after testifying. Case perceptions included a dichotomous responsibility judgment, strength of responsibility ratings, and a cognitive dissonance measure. Results: Concurrent testimony did not eliminate adversarial allegiance. For perceptions of responsibility, there was no significant difference between adversarial and concurrent experts (ηp2 = .001), nor any change in participants’ ratings over time (ηp2 = .03); however, prosecution experts – across testimony types – rated the defendant as significantly more responsible compared to defense experts (ηp2 = .80). Concurrent and adversarial experts did not differ in the information provided in their expert reports and minimally differed in testimony content.Conclusions: Partisan experts showed adversarial allegiance regardless of expert testimony method, and we observed no attenuation of this bias over the course of the evaluation.