Objective To evaluate the effectiveness of small‐group nurse‐administered cognitive behavioural therapy for insomnia (CBTI) as an early intervention of mood disorders with comorbid insomnia. Methods A total of 200 patients with first‐episode depressive or bipolar disorders and comorbid insomnia were randomized in a ratio of 1:1 to receiving 4‐session CBTI or not in a routine psychiatric care setting. Primary outcome was Insomnia Severity Index. Secondary outcomes included response and remission status; daytime symptomatology and quality of life; medication burden; sleep‐related cognitions and behaviours; and the credibility, satisfaction, adherence and adverse events of CBTI. Assessments were conducted at baseline, 3, 6, and 12‐month. Results Only a significant time‐effect but no group‐by‐time interaction was found in the primary outcome. Several secondary outcomes had significantly greater improvements in CBTI group, including higher depression remission at 12‐month (59.7% vs. 37.9%, χ2 = 6.57, p = .01), lower anxiolytic use at 3‐month (18.1% vs. 33.3%, χ2 = 4.72, p = .03) and 12‐month (12.5% vs. 25.8%, χ2 = 3.26, p = .047), and lesser sleep‐related dysfunctional cognitions at 3 and 6‐month (mixed‐effects model, F = 5.12, p = .001 and .03, respectively). Depression remission rate was 28.6%, 40.3%, and 59.7% at 3, 6, and 12‐month, respectively in CBTI group and 28.4%, 31.1%, and 37.9%, respectively in no CBTI group. Conclusion CBTI may be a useful early intervention to enhance depression remission and reduce medication burden in patients with first‐episode depressive disorder and comorbid insomnia.