Objective: This randomized controlled trial evaluated the efficacy of an Internet-based intervention, which aimed to improve recovery from work-related strain in teachers with sleeping problems and work-related rumination. In addition, mechanisms of change were also investigated. Methods: A sample of 128 teachers with elevated symptoms of insomnia (Insomnia Severity Index [ISI] ≥ 15) and work-related rumination (Cognitive Irritation Scale ≥ 15) was assigned to either an Internet-based recovery training (intervention condition [IC]) or to a waitlist control condition (CC). The IC consisted of 6 Internet-based sessions that aimed to promote healthy restorative behavior. Self-report data were assessed at baseline and again after 8 weeks. Additionally, a sleep diary was used starting 1 week before baseline and ending 1 week after postassessment. The primary outcome was insomnia severity. Secondary outcomes included perseverative cognitions (i.e., work-related rumination and worrying), a range of recovery measures and depression. An extended 6-month follow-up was assessed in the IC only. A serial multiple mediator analysis was carried out to investigate mechanisms of change. Results: IC participants displayed a significantly greater reduction in insomnia severity (d = 1.37, 95% confidence interval: 0.99–1.77) than did participants of the CC. The IC was also superior with regard to changes in all investigated secondary outcomes. Effects were maintained until a naturalistic 6-month follow-up. Effects on insomnia severity were mediated by both a reduction in perseverative cognitions and sleep effort. Additionally, a greater increase in number of recovery activities per week was found to be associated with lower perseverative cognitions that in turn led to a greater reduction in insomnia severity. Conclusions: This study provides evidence for the efficacy of an unguided, Internet-based occupational recovery training and provided first evidence for a number of assumed mechanisms of change.