Objective To assess the efficacy and safety of mirtazapine and fluoxetine on depression. Methods We searched The Cochrane Library (Issue 2, 2009), MEDLINE (1980 to Dec.2008), EMbase (1980 to Dec.2008), CBM (1980 to Dec.2008), VIP (1980 to Dec.2008), CNKI (1980 to Dec.2008) and Wanfang database (1980 to Dec.2008) to search randomized controlled trials (RCTs) comparing mirtazapine with fluoxetine for depression. The quality of the included trials was assessed and meta-analysis was conducted by RevMan 5.0 software. Results Five RCTs involving 695 patients were included. The results of meta-analyses showed that: (1) After one-week treatment and two-week treatment, effectiveness of the mirtazapine group was significantly higher than that of the fluoxetine group [one-week treatment: RR=2.00, 95%CI (1.30, 3.10), P=0.002; two-week treatment: RR=1.49, 95%CI (1.08, 2.06), P=0.02]. But after six-week treatment, there was no significant difference of the efficacy between the mirtazapine and the fluoxetine groups with RR=1.21 and 95%CI 0.89 to 1.63 (P=0.23), and the cure rate of the mirtazapine group was a little higher than that of the fluoxetine group with RR=1.40 and 95%CI 1.09 to 1.80 (P=0.009). (2) Side reaction: The somnolence and weight gain rates of the mirtazapine group was higher than those of the fluoxetine group with RR=1.78 and 95%CI 1.18 to 2.70 (P=0.006) and RR=5.91 and 95%CI 2.21 to 15.83 (P=0.000 4). But fluoxetine more easily induced nausea and insomnia with RR=0.47 and 95%CI 0.31 to 0.71 (P=0.000 3); RR=0.39 and 95%CI 0.17 to 0.89 (P=0.03) than mirtazapine. And the other common side reactions were all not significantly different between the two groups (Pgt;0.05). Conclusion Mirtazapine is more effective than fluoxetine and works faster. Mirtazapine could more easily induce somnolence and weight gain, and is with lower nausea or insomnia rate when compared with fluoxetine. And there is no significant difference in comparison of other common side reactions.