Objectives To determine the efficacy and safety of blue-light therapy in seasonal and non-seasonal major depressive disorder (MDD), by comparison to active and inactive control conditions. Methods We searched Web of Science, EMBASE, Medline, PsycInfo, and Clinicaltrials.gov through January 17, 2022, for randomized controlled trials (RCTs) using search terms for blue/blue-enhanced, light therapy, and depression/seasonal affective disorder. Two independent reviewers extracted data. The primary outcome was the difference in endpoint scores on the Structured Interview Guide for the Hamilton Depression Rating Scale - Seasonal Affective Disorder (SIGH-SAD) or the Structured Interview Guide for the Hamilton Depression Rating Scale with Atypical Depression Supplement (SIGH-ADS) between blue light and comparison conditions. Secondary outcomes were response (≥ 50% improvement from baseline to endpoint on a depression scale) and remission rates (endpoint score in the remission range). Results Of 582 articles retrieved, we included nine RCTs ( n = 347 participants) assessing blue-light therapy. Seven studies had participants with seasonal MDD and two studies included participants with non-seasonal MDD. Four studies compared blue light to an inactive light condition (efficacy studies), and five studies compared it to an active condition (comparison studies). For the primary outcome, a meta-analysis with random-effects models found no evidence for the efficacy of blue-light conditions compared to inactive conditions (mean difference [MD] = 2.43; 95% confidence interval [CI], −1.28 to 6.14, P = 0.20); however, blue-light also showed no differences compared to active conditions (MD = −0.11; 95% CI, −2.38 to 2.16, P = 0.93). There were no significant differences in response and remission rates between blue-light conditions and inactive or active light conditions. Blue-light therapy was overall well-tolerated. Conclusions The efficacy of blue-light therapy in the treatment of seasonal and non-seasonal MDD remains unproven. Future trials should be of longer duration, include larger sample sizes, and attempt to better standardize the parameters of light therapy.