Objective To evaluate the effect of nursing interventions on the medical behavior of diabetic patients. Methods We searched CMB (1978 to September 2007), CNKI (1994 to September 2007), VIP (1989 to September 2007) and Papers on Academic Conference of China (1989 to September 2007), and hand-searched relevant journals. We identified randomized controlled trials (RCTs) and quasi-RCTs of nursing interventions plus routine therapy versus simple routine therapy for diabetic patients. The methodological quality of included studies was assessed and data was extracted and analyzed by using The Cochrane Collaboration’s RevMan 4.2.10 software. Results In total, 10 RCTs and 2 quasi-RCTs were included. Two of these studies described blinding. The quality of the included trials was low. Meta-analyses showed that nursing interventions significantly improved medication compliance (RR 1.33, 95%CI 1.21 to 1.45), medical behaviors of physical therapy (RR 1.48, 95%CI 1.30 to 1.69), diet control (RR 1.54, 95%CI 1.39 to 1.71), blood glucose monitoring (RR 1.56, 95%CI 1.41 to 1.72) and regular inspection (RR 1.66, 95%CI 1.41 to 1.96) in patients with diabetes. Conclusion Nursing interventions can improve the medical behavior of diabetic patients, increase patient-controlled capacity and enhance awareness of diabetes. It is also beneficial for blood glucose control and stability.