While most psychotherapy methods are about equally effective, it is unclear if (1) different methods of psychotherapy differ in the severity of negative effects, (2) negative effects impact outcome and (3) this impact is moderated by psychotherapy method. We analyzed data from an observational study of 141 patients from a day treatment program for depression. Based on shared decision-making, patients were treated with either Cognitive Behavioral Analysis System of Psychotherapy (CBASP) or Metacognitive Therapy (MCT). Negative effects were assessed with the Negative Effects Questionnaire (NEQ), severity of depressive symptoms with the Quick Inventory of Depressive Symptomatology (QIDS-SR). Treatment groups were propensity score matched to account for baseline differences. Severity of negative effects did not differ between CBASP and MCT. The association between negative effects and outcome was moderated by treatment method. For patients treated with CBASP, negative effects were associated with outcome: those experiencing the lowest severity of negative effects had the greatest improvement in symptoms during treatment. Treatments were equally tolerated but differed in their association between negative effects and outcome. Results need to be considered with caution due to the considerable drop-out rate during the follow-up period and the non-controlled nature of our study.