OBJECTIVE: To examine the efficacy and safety of hypnotherapy for insomnia as compared to placebo, pharmacological or non-pharmacological intervention, or no treatment. METHODS: A systematic search on major electronic databases was conducted up until March 2014. Inclusion criteria are: (1) randomized controlled trials (RCTs) or quasi-RCTs; (2) intervention targeted at improving sleep; (3) hypnosis as an intervention; and (4) English language articles. Sleep diary variable is the primary outcome measure. RESULTS: Six RCTs of hypnotherapy and seven on autogenic training or guided imagery, comprising 502 subjects, were included. Eleven of the 13 studies had low methodological quality, as indicated by a modified Jadad score below 3, and high risks of bias in blinding and design of the control interventions. No adverse events related to hypnosis were reported, though seldom investigated. Meta-analyses found hypnotherapy significantly shortened sleep latency compared to waitlist (standardized mean difference, SMD=-0.88, 95% confidence interval (CI): -1.56, -0.19, P=0.01, I(2)=15%), but no difference compared to sham intervention (SMD: -1.08, 95% CI: -3.15, 0.09, P=0.31, I(2)=90%). Similar results were found for autogenic training or guided imagery (SMD with waitlist=-1.16, 95% CI: -1.92, -0.40, P=0.003, I(2)=0%; SMD with sham intervention=-0.50, 95% CI: -1.19, 0.19, P=0.15, I(2)=0%). CONCLUSIONS: Generalizability of the positive results is doubtful due to the relatively small sample size and methodological limitations. Future studies with larger sample size and better study design and methodology are called for.