Objective: Cognitive therapy (CT) and behavior therapy (BT) are both effective for insomnia but are expected to work via different pathways. Empirically, little is known about their symptom-specific effects. Method: This was a secondary analysis of a randomized controlled trial of online treatment for insomnia disorder (N = 219, 72.9% female, mean age = 52.5 years, SD = 13.9). Participants were randomized to CT (n = 72), BT (n = 73), or wait-list (n = 74). Network Intervention Analysis was used to investigate the symptom-specific treatment effects of CT and BT throughout treatment (wait-list was excluded from the current study). The networks included the Insomnia Severity Index items and the sleep diary-based sleep efficiency and were estimated biweekly from Week 0 until Week 10. Results: Participants in the BT condition showed symptom-specific effects compared to CT on "sleep efficiency" (Week 4-8, post-test), "difficulty maintaining sleep" (Week 4), and "dissatisfaction with sleep" (post-test). Participants in the CT showed symptom-specific effects compared to BT on "interference with daily functioning" (Week 8, post-test), "difficulty initiating sleep", "early morning awakenings," and "worry about sleep" (all post-test). Conclusions: This is the first study that observed specific differential treatment effects for BT and CT throughout the course of their treatment. These effects were more pronounced for BT than for CT and were in line with the theoretical background of these treatments. We think the embedment of the theoretical background of CT and BT in empirical data is of major importance to guide further treatment development. (PsycInfo Database Record (c) 2021 APA, all rights reserved).