Objective To evaluate the safety of mycophenolate mofetil (MMF) versus azathioprine (AZA) for rejection after renal transplantation. Methods We searched MEDLINE (1966 to Jun. 2004), EMBASE (1984 to Jun. 2004), The Cochrane Library (Issue 2, 2004) and Chinese Biomedical Database (CBM, 1979 to Jun. 2004). Randomized controlled trials (RCTs) comparing MMF with AZA for rejection after renal transplantation were included. The quality of included studies such as randomization, blinding, allocation concealment was evaluated and meta-analysis was performed using RevMan 4.1.1 software. Results Twenty-Four RCTs comparing MMF (2 g/day or 3 g/day) with AZA for rejection after renal transplantation were identified. The digest system morbidity of MMF group was higher than that of AZA group. The incidence of vomiting, bellyache and diarrhea of MMF 3 g/day group was statistical by higher than that of AZA group (P＜0.05). The cytom egalovirus (CMV) infection morbidity of MMF 3 g/day group during 6 months, 1 year and 2 years follow-up was higher than AZA group with statistical difference, but for MMF 2 g/day group, this difference was only seen during 1 year follow-up. Leukopenia incidence of MMF 3g/day group was higher than AZA group with statistical difference, but this difference was not seen in MMF 2 g/day group. Thrombocytopenia incidence of MMF 3 g/day group was lower than AZA group with statistical difference. For skin carcinoma morbidity, no statistical difference was found among MMF 3 g/day, MMF 2 g/day and AZA groups. Conclusions Compared with AZA, MMF represents higher digest system side-effects incidence, higher morbidity of leucopenia and CMV infection and lower incidence of thrombocytopenia. The dose-response relationship of adverse drug reaction is found.