OBJECTIVE: To determine the efficacy of acupuncture in the treatment of depression by performing systematical reviews and meta-analysis from 20 articles of clinical, randomized, controlled studies. DATA SOURCES: A computer-based online search of PubMed Database, Chinese Biological Medicine Database (VIP database), China Knowledge Database, and Wanfang Database was performed to search articles published between January 1999 and September 2009. DATA SELECTION: Inclusion criteria: articles published in official academic journals; articles with definite diagnostic criteria and acknowledgment; articles involving subjects with depression or depressive disorders, with no limitations to age, sex, or source; articles involving acupuncture-related therapies (including acupuncture needle, electro-acupuncture, moxibustion, abdominal acupuncture, etc.) as the primary means of intervention; articles of randomized controlled trials or involving "randomized controlled", "randomized grouping", or "random". Jadad scores were utilized to evaluate quality of included articles. Review Manage 4.2.7 software was used to evaluate efficacy of acupuncture in treating depression, and Hamilton Depression Rating Scale (HAMD) score, Self-Rating Depression Scale (SDS) score, and meta-analysis were used to evaluate clinical efficacy. Meta-analysis of a fixed effect model was performed when P > 0.05 or P > 0.01 in the test for heterogeneity. Meta-analysis of the random effect model was performed when P <=. 0.05. MAIN OUTCOME MEASURES: Efficacy rate, HAMD scores, and SDS scores. RESULTS: Quality evaluation of the 20 selected clinical, randomized, controlled trials demonstrated baseline equilibrium and randomness, although a considerable number of randomized studies simply mentioned "random" without describing a specific method. Therefore, the reliability of randomness was relatively low. An envelope to hide the information was not used, and proper blinding was difficult to implement due to the innate features of acupuncture. The follow-up rate was low, and reasons for dropout were not explored. Therefore, it was not possible to objectively evaluate the study results. In addition, the study did not estimate sample size; it was unclear whether the sample was truly representative of general characteristics. Meta-analysis showed that acupuncture monotherapy exhibited similar efficacy to Western medicine for treating depression: combined effect size OR = 1.66, 95% CI: 0.59-4.65, combined effect value test: Z = 0.97, P = 0.33. Comparison of the efficacy of acupuncture combination therapy with Western medicine demonstrated a combined effect size OR = 2.46, 95% CI: 1.64-3.71 and combined effect of the value of Z = 4.32 (P < 0.01), implying that there was insufficient evidence to prove a superior effect of acupuncture combination therapy over Western medicine. In terms of HAMD and SDS scores, the weighted mean difference was -2.01 and -8.68, respectively, with 95% CI: -3.48 to 0.53, -11.21 to -6.16 (P < 0.01), suggesting that acupuncture efficacy was significantly better than the control group. CONCLUSION: The efficacy of acupuncture monotherapy was similar to Western medicine for treating depression. However, efficacy of combination therapy remains uncertain due to inadequate study design (implementation of randomization and blinding, sample size estimation, and data processing methods).