Objective To determine the efficacy and safety of Amplatzer transcatheter closure and surgical closure for ostium secundum atrial septal defects. Methods MEDLINE (1966-July 2006), EMBASE (1966-July 2006), The Cochrane Library (Issue 2, 2006) and CBMdisc (1979-July 2006) were searched for randomized controlled trials or non-randomized controlled trials. Data were extracted by two reviewers using a specially designed extraction form. The quality of included trials was critically assessed. The Cochrane Collaboration’s RevMan 4.2 software was used for data analysis. Results Sixteen non-RCTs involving 2 043 patients were included. No deaths were reported in 14 trials. The other two trials reported one death respectively in the surgical group. Meta-analysis of 12 trials involving 1 722 patients showed that the procedure success rate for the Amplatzer group was lower than that for the surgical group [WMD:0.95, 95%CI (0.92,0.98)]. Similar results were also found in the complete closure rate in 24-hour follow-up [6 trials involving 1 106 patients, WMD:0.96, 95%CI (0.92 to 1.00)], the complication rate [16 trials involving 1 971 patients, WMD:0.27, 95%CI(0.21 to 0.35)] and the transfusion rate [14 trials involving 1 807 patients, WMD:0.03, 95%CI(0.02 to 0.06)]. Conclusions The success rate for Amplatzer device closure of ASD is lower than that of surgical repair. However, the complication rate, length of hospital stay and transfusion rate are lower or shorter for Amplatzer device closure than for surgical repair. Based on appropriate patient selection, Amplatzer closure of ASD is a safe and effective alternative to surgical repair.