Although Stress Management Training (SMT) appears to result in modest benefits for participants, it is unclear if these benefits are due to non-specific factors or specific skills related interventions. SMT with health care staff was delivered in six weekly sessions, with each session devoted to specific tasks and goals. Using a session impact methodology developed in psychotherapy research, participants rated each session in terms of depth, smoothness, post-session mood and 12 specific impacts, including task and interpersonal positive impacts, and problematic, negative impacts. As a validity check, SMT session ratings were compared with session ratings from psychotherapy sessions. As expected, SMT sessions resembled sessions of cognitive behavioural psychotherapy and differed from sessions of psychodynamic/interpersonal therapy. Impact ratings of SMT sessions were of three types: those which did not differ between sessions; those which appeared to reflect non-specific and group processes—which showed a significant linear trend over time; and those which reflected specific session content where no linear trend was detected but one or more sessions differed significantly from others. As in psychotherapy research, session impact ratings are a promising method of identifying specific mechanisms of change in worksite intervention studies.