Objective To compare the efficacy of plating versus intramedullary nailing in the treatment of adult humeral shaft fracture. Methods We identified eligible studies in PubMed (1950 to September 2007), MEDLINE (1950 to September 2007), OVID CINAHL (1950 to September 2007), OVID EBM (3rd Quarter 2007), CBMdisk (1978 to June 2007) and CNKI (1981 to June 2007). We also handsearched several Chinese orthopedic journals. Data were extracted and evaluated by two reviewers independently. Randomized controlled trials (RCTs) comparing plating versus intramedullary nailing for humeral shaft fracture in adults were included and the quality of these trials was critically assessed. Data analyses were conducted with Stata 10.0. Results Six RCTs involving 425 patients were included, among which the statistical heterogeneity was not significant (Pgt;0.1). Cumulative meta-analyses showed that intramedullary nailing might increase the re-operation rate in studies conducted before the year of 2000 (OR=0.39, 95%CI 0.17 to 0.90, P=0.03), but the difference was not significant in studies conducted after 2000 (OR=0.54, 95%CI 0.27 to 1.08, P=0.08). Intramedullary nailing might increase the incidence of shoulder impingement compared with plating (OR=0.13, 95%CI 0.03 to 0.65, P=0.01). The rates of non-union, deep infection, iatrogenic radial nerve injury and internal fixation failure were similar between plating and intramedullary nailing. Meta-analyses were not conducted for union time, operation time and bleeding (transfusion) volume, because the relevant data were not available from the included trials. Conclusion Intramedullary nailing may increase the incidence of shoulder impingement. The rates of re-operation, non-union, deep infection, iatrogenic radial nerve injury and internal fixation failure are similar between plating and intramedullary nailing. Further well-designed and large-scale randomized controlled trials are required to determine the effects of plating and intramedullary nailing on these outcomes.