BACKGROUND: The study aimed to determine the efficacy of melatonin supplementation for sleep disturbances in patients with traumatic brain injury (TBI). METHODS: This is a randomised double-blind placebo-controlled two-period two-treatment (melatonin and placebo) crossover study. Outpatients were recruited from Epworth and Austin Hospitals Melbourne, Australia. They had mild to severe TBI (n = 33) reporting sleep disturbances post-injury (mean age 37 years, standard deviation 11 years; 67% men). They were given prolonged-release melatonin formulation (2 mg; Circadin®) and placebo capsules for 4 weeks each in a counterbalanced fashion separated by a 48-hour washout period. Treatment was taken nightly 2 hours before bedtime. Serious adverse events and side-effects were monitored. RESULTS: Melatonin supplementation significantly reduced global Pittsburgh Sleep Quality Index scores relative to placebo, indicating improved sleep quality [melatonin 7.68 vs. placebo 9.47, original score units; difference -1.79; 95% confidence interval (CI), -2.70 to -0.88; p ≤ 0.0001]. Melatonin had no effect on sleep onset latency (melatonin 1.37 vs. placebo 1.42, log units; difference -0.05; 95% CI, -0.14 to 0.03; p = 0.23). With respect to the secondary outcomes, melatonin supplementation increased sleep efficiency on actigraphy, and vitality and mental health on the SF-36 v1 questionnaire (p ≤ 0.05 for each). Melatonin decreased anxiety on the Hospital Anxiety Depression Scale and fatigue on the Fatigue Severity Scale (p ≤ 0.05 for both), but had no significant effect on daytime sleepiness on the Epworth Sleepiness Scale (p = 0.15). No serious adverse events were reported. CONCLUSIONS: Melatonin supplementation over a 4-week period is effective and safe in improving subjective sleep quality as well as some aspects of objective sleep quality in patients with TBI. TRIAL REGISTRATION: Identifier: 12611000734965; Prospectively registered on 13 July 2011.