OBJECTIVES: To assess the effectiveness and safety of  nevirapine (NVP) for preventing mother-to-child transmission (MTCT) of HIV. METHODS: We searched the databases of PubMed, EMBASE, CINAHL, AID Search, AIDS info and three Chinese Databases (CBM, CNKI, VIP) from the beginning of the databases to December 30, 2007. The Cochrane Library (2007 issue 4) was also searched. Randomized Controlled Clinical Trials (RCT) assessing the effects of NVP for preventing MTCT were included. Trial selection, quality assessment and data extraction were done by two reviewers independently, with confirmation by cross-checking the information. Meta-analyses were conducted by using RevMan 4.2.9 software. RESULTS: Eight trials were included. Of them, three trials with 4059 participants compared the nevirapine with standard antiretroviral therapy (ART). One trial(Thai-PHPT, 1844 participants) demonstrated that when nevirapine was given to mothers already receiving standard antiretroviral therapy with ZDV, both the NVP-NVP and NVP-Placebo decreased the risk of transmission by 91% or 82% than placebo-placebo group. However, due to the use of highly active ART (HAART), there appeared to be no additional advantage(p>0.05). Two trials(Thai-PHPT2 and Mashi-part1, 2553 participants) showed that in the setting of short course zidovudine, the effectiveness and the incidences of any adverse events in infant or maternal single dose nevirapine(sdNVP) group were similar with in the group where both mother and infant received sdNVP. Meta-analysis of 3 RCT (1753 participants) showed that that sdNVP given to mothers and babies was more effective than an intrapartum and post-partum regimen of zidovudine (decreased the risk by 44%-65%). The incidences of stillbirth, infant and maternal mortality, low birth weigh, any adverse events in mothers and children were similar between the two groups (p>0.05). Three RCTs (3153 participants) compared the effect of NVP plus ZDV with NVP alone. When the mother and infant did not receive any ART or NVP, ZDV appeared additional advantage either in HIV infection rate or cumulative rate (RR=0.63,95%CI=0.42-0.96, p=0.03; RR=0.73, 95%CI=0.56-0.96, p=0.02). However, when the mother and infant received intrapartum NVP plus ZDV or sdNVP, there appeared no additional advantage (p>0.05). The incidences of stillbirth, grade 3 or greater ADR and average birth weigh were similar between the two groups (p>0.05). CONCLUSIONS: Single dose of NVP is more effective than short course and ultra-short course ZDV with less expensive and easier to use. It is recommended as minimum regime for preventing MTCT of HIV by WHO with the support of evidence. When highly active ART (HAART) is adopted to prevent the peripartum transmission of HIV, sdNVP(both to mother and to infant) appears no additional advantage.