Objective To systematically evaluate effectiveness and safety of nerve block therapy for neck pain. Methods Databases including CENTRAL, PubMed, Ovid, ISI, EBSCO, CBM and CNKI were searched from the date of their establishment to November 2011, and relevant references were also retrieved manually to collect both domestic and abroad randomized controlled trials (RCTs) about nerve block therapy for neck pain. According to the inclusion and exclusion criteria, two researchers independently screened literature, extracted data, and assessed the quality of the included studies. Then the meta-analysis was conducted using RevMan 5.0 software. Results A total of 10 studies involving 625 participants were included. The results of qualitative analysis showed that: a) The short-term effectiveness of the nerve block therapy group was markedly superior to the placebo group, the cognitive therapy group and the transcutaneous electric nerve stimulation (TENS) group; and b) The short-term effectiveness of the combined nerve block therapy was markedly superior to the single nerve block therapy. The results of meta-analysis demonstrated that: a) There was no significant difference between the greater occipital nerve (GON) block group and the C2/3 nerve block group in neither short-term (SMD=−0.13, 95%CI −0.58 to 0.32, P=0.58) nor medium-term effectiveness (SMD=−0.01, 95%CI −0.46 to 0.44, P=0.98); and b): There was no significant difference between the injection with steroids group and the injection without steroids group in both short-term (SMD=0.16, 95%CI −0.13 to 0.44, P=0.28) and long-term effectiveness (SMD=0.27, 95%CI −0.02 to 0.55, P=0.07). Conclusion Current evidence shows nerve block therapy for neck pain is safe and especially good in short-term effectiveness. The combined nerve block therapy is probably more effective, but the effectiveness is not obviously improved by injection with or without steroids, and by different block methods. Due to the limitation of quality, quantity and total sample size of the included studies, this conclusion still needs to be proved by conducting more high quality and large scale studies.