INTRODUCTION: Little is known on what pretreatment patient characteristics the outcome of Cognitive Therapy (CT) and Behavioral Therapy (BT) for insomnia disorder depends on. Identifying for whom treatment is most useful is an essential step toward treatment optimization and personalized care. Therefore, the purpose with this investigation was to examine both theory-driven constructs and insomnia-associated clinical variables as potential predictors and moderators of outcome in CT and BT. MATERIALS AND METHODS: One hundred and forty-five participants diagnosed with insomnia disorder were randomized to 10 weekly internet-delivered modules of CT or BT with 15 minutes of weekly telephone support. General clinical predictors and theory-driven moderators (cognitive and behavioral processes), assessed in a former RCT, were analyzed using multiple linear regression with insomnia severity as the outcome. RESULTS: Bedtime variability and early morning waketime interacted with treatment and indicated that lower bedtime variability and early morning waketime were associated with a higher effect for CT, whereas the opposite was true for BT. Waketime after sleep onset, insomnia severity index, and sleep efficiency emerged as predictors, indicating prognostic value of treatment outcome. CONCLUSIONS: Five constructs provided predictive values in the outcome of cognitive therapy and behavior therapy. The moderator findings are in line with the theoretical models of CT and BT and may have implications for future research and clinical practice of CBT-I, should they be replicated. Clinically, this could implicate the ability to match therapy to patient features in order to optimize outcomes.