OBJECTIVES: Meta-analyses (MAs) of randomized controlled trials (RCTs) are considered to provide the highest level of evidence. There has been a substantial increase in the number of published MAs, but of unknown quality. Therefore, we determined the quality of reporting in MAs of RCTs of pharmacotherapy used in the treatment of major depressive disorder (MDD) in adults (18-65) without comorbidities. We also examined trends over time. METHODS: MEDLINE, EMBASE, Healthstar, Psychlit and Cochrane databases were searched from 1980-2002 by 4 independent reviewers for MAs of RCTs reporting efficacy rates. Articles meeting inclusion criteria were blinded. Interrater reliability (Kappa) and test-retest between 4 raters was evaluated using 4 articles. MA quality was assessed using the QUORUM checklist comprising 6 sections (Title, Abstract, Introduction, Methods, Results, Discussion) with 18 questions in total, each awarded 1 point. Descriptive statistics were used to calculate overall MA quality. Time trends were evaluated by calculating Spearman's rho. RESULTS: Fifty-nine articles were identified, 26 were excluded (comorbidities (12), inappropriate compartor (7), no RCT (1), article not available (6)); 33 were included. Initial kappa was 0.81 (p<0.05). After resolution of disagreements through consensus, the test-retest reliability was significant (kappa=0.89; p<0.05). The average overall quality score was 50.3% (SD=15.6%, range=16.7%-88.9%, median=50.0%). The overall score for Titles was very poor (21%), Abstracts (39%) and Methods (49%) had poor scores, while the overall Results score was minimally acceptable (55%). Good quality scores were found for the Introduction (91%) and Discussion (100%). No time trends were identified (rho=0.08; p=0.68). CONCLUSIONS: Although quality guidelines have been published, the average quality of published MAs of antidepressants is barely acceptable (50.3%). There is a need for adherence to standardized reporting and quality guidelines.