Background: Insomnia has become a major public health concern. Aims: The study examined the efficacy of a web-based unguided self-help programme with automated feedback. The programme was based on cognitive behaviour therapy for insomnia (CBT-I). The investigation particularly focused on factors that contribute to the maintenance of insomnia and tested whether treatment effects were stable over a period of 12 months. Method: Fifty-six participants were randomly assigned either to web-based CBT-I or to the waiting-list control group. Included measures assessed insomnia severity, sleep-related cognitions, safety behaviours, depression, anxiety and somatization. In the intervention group, a sleep diary was used to assess sleep continuity parameters, sleep quality and daytime performance. Results: Large between- and within-group effect sizes (d = 1.79, d = 1.59) for insomnia severity were found. The treatment group effect remained stable over the period of 12 months. Further, sleep-related cognitions, safety behaviours, depression and somatization significantly decreased in the treatment group compared with the control group. On all sleep diary parameters, medium to large effects were revealed within the treatment group. Anxiety did not decrease significantly from pre- to post-assessment. For all measures except somatization and anxiety significant within-group effects were found at 12-month follow-up assessment indicating long-lasting effects. Conclusions: This study adds evidence to the literature on unguided online interventions for insomnia, and indicates that online CBT-I can have substantial long-term effects on relevant sleep-related outcome parameters. Moreover, the results indicate that sleep-related cognitions and safety behaviour can be successfully altered with an unguided CBT-I intervention.