Background In the psychological therapies, patient outcomes are not always positive. Some patients leave therapy prematurely (dropout), while others experience deterioration in their psychological well‐being. Methods The sample for dropout comprised patients (n = 10 521) seen by 85 therapists, who attended at least the initial session of one‐to‐one therapy and completed a Clinical Outcomes in Routine Evaluation‐Outcome Measure (CORE‐OM) at pre‐treatment. The subsample for patient deterioration comprised patients (n = 6405) seen by the same 85 therapists but who attended two or more sessions, completed therapy and returned a CORE‐OM at pre‐treatment and post‐treatment. Multilevel modelling was used to estimate the extent of therapist effects for both outcomes after controlling for patient characteristics. Results Therapist effects accounted for 12.6% of dropout variance and 10.1% of deterioration variance. Dropout rates for therapists ranged from 1.2% to 73.2%, while rates of deterioration ranged from 0% to 15.4%. There was no significant correlation between therapist dropout rate and deterioration rate (Spearman's rho = 0.07, p = 0.52). Conclusions The methods provide a reliable means for identifying therapists who return consistently poorer rates of patient dropout and deterioration compared with their peers. The variability between therapists and the identification of patient risk factors as significant predictors has implications for the delivery of safe psychological therapy services. Copyright © 2016 John Wiley & Sons, Ltd. Key Practitioner Message Therapists play an important role in contributing to patient dropout and deterioration, irrespective of case mix. Therapist effects on patient dropout and deterioration appear to act independently. Being unemployed as a patient was the strongest predictor of both dropout and deterioration. Patient risk to self or others was also an important predictor.