While cognitive-behavioral therapy (CBT) is the most well-established treatment for binge-eating disorder (BED), little is known about process factors influencing its outcome. The present study sought to explore the assessment of therapist adherence, its course over treatment, and its associations with patient and therapist characteristics, and the therapeutic alliance. In a prospective multicenter randomized-controlled trial comparing CBT to internet-based guided self-help (INTERBED-study), therapist adherence using the newly developed Adherence Control Form (ACF) was determined by trained raters in randomly selected 418 audio-taped CBT sessions of 89 patients (25% of all sessions). Observer-rated therapeutic alliance, interview-based and self-reported patient and therapist characteristics were assessed. Three-level multilevel modeling was applied. The ACF showed adequate psychometric properties. Therapist adherence was excellent. While significant between-therapist variability in therapist adherence was found, within-therapist variability was non-significant. Patient and therapist characteristics did not predict the therapist adherence. The therapist adherence positively predicted the therapeutic alliance. The ACF demonstrated its utility to assess therapist adherence in CBT for BED. The excellent levels of therapist adherence point to the internal validity of the CBT within the INTERBED-study serving as a prerequisite for empirical comparisons between treatments. Variability between therapists should be addressed in therapist trainings and dissemination trials.

While cognitive-behavioral therapy (CBT) is the most well-established treatment for binge-eating disorder (BED), little is known about process factors influencing its outcome. The present study sought to explore the assessment of therapist adherence, its course over treatment, and its associations with patient and therapist characteristics, and the therapeutic alliance. In a prospective multicenter randomized-controlled trial comparing CBT to internet-based guided self-help (INTERBED-study), therapist adherence using the newly developed Adherence Control Form (ACF) was determined by trained raters in randomly selected 418 audio-taped CBT sessions of 89 patients (25% of all sessions). Observer-rated therapeutic alliance, interview-based and self-reported patient and therapist characteristics were assessed. Three-level multilevel modeling was applied. The ACF showed adequate psychometric properties. Therapist adherence was excellent. While significant between-therapist variability in therapist adherence was found, within-therapist variability was non-significant. Patient and therapist characteristics did not predict the therapist adherence. The therapist adherence positively predicted the therapeutic alliance. The ACF demonstrated its utility to assess therapist adherence in CBT for BED. The excellent levels of therapist adherence point to the internal validity of the CBT within the INTERBED-study serving as a prerequisite for empirical comparisons between treatments. Variability between therapists should be addressed in therapist trainings and dissemination trials.