Roughly 25% of patients with insomnia do not respond to the current first line of treatment: Cognitive Behavioural Therapy for Insomnia (CBT-I). Evidence for Acceptance and Commitment Therapy for Insomnia (ACT-I) as an alternative treatment is still limited. This RCT aims to investigate ACT-I’s efficacy as a stand-alone treatment in improving insomnia, while also investigating the role of psychological (in)flexibility as mediator. Ninety-three participants were randomized to a 5-session ACT-I intervention or a waitlist control group. Insomnia severity, psychological flexibility and psychological inflexibility were measured before, during and after treatment. In line with the hypotheses, participants in the ACT-I group displayed larger reductions in insomnia severity between the pre- and post-treatment assessments and were more likely to achieve a clinically significant change (35.7%, n = 15) or full remission (21.4%, n = 9) than participants in the waitlist group (6.7%, n = 3; 2.2%, n = 1, respectively). Contradicting our hypotheses, early changes in psychological flexibility and inflexibility between the pre- and mid-treatment assessments did not mediate the treatment effect on insomnia severity. However, changes between the pre- and post-treatment assessments in psychological inflexibility did partially mediate the treatment effect. The findings support ACT-I as effective insomnia treatment, albeit with lower response and remission rates than previously seen for CBT-I. We found limited support for the role of psychological inflexibility as mediator. ACT-I poses a promising alternative treatment for insomnia patients who do not respond to CBT-I.