OBJECTIVES: To systematically evaluate and assess the effectiveness of antidepressants in treating depressed patients after stroke. METHODS: A systematic review with Meta analysis of double-blinded randomized clinical trials (RCTs) about the treatment of depression in the patients with stroke. Data were collected from the published RCTs, which indexed in the Medline from 1960 to 2005. Review was only limited in the English articles. Outcome measurements: 1) The proportion of treatment responders whose depression scores improved by >50% was compared. 2) The depression rating scores at the beginning, middle, and end of treatments . RESULTS: A total of 313 patients from 5 RCTs fulfilling the inclusion criteria were identified for this study. The percentages of responders in treatment and control group were 57.9% and 44.4%, respectively (Risk Difference, RD=0.17, 95%CI from -0.04 to 0.38, Z=1.61, P=0.11). In term of depression scores, prior to any treatments, the overall depression scores in the patients of treatment groups were a little bit higher (more depressed) than the patients in the control groups, but the difference was not statistically significant (Weighted Mean Difference, WMD=1.01,Z=1.59, P=0.11). After the treatments, patients in the treatment groups became statistically less depressed than the patients in the control groups in term of lower depression scores (WMD=-1.88, Z=2.16, P=0.03). In addition, we found that the overall WMD in term of depression scores between two groups was 0.60 (Z=0.87, p=0.39) at the beginning of treatment, and became -1.85 (Z=2.12, P=0.03) in the middle of treatment, then further decreased to -3.06 (Z=2.92, p=0.004) at the end of treatment. CONCLUSION: The results indicated that antidepressant treatments were effective in the patients after stroke in term of reducing the depression scores and treatment effects were time dependent. In addition, more efforts should be guaranteed to study the cerebrovascular effects of antidepressants in patients after stroke.