BACKGROUND: The advent of highly active antiretroviral therapy (ART) has reduced the morbidity and mortality due to HIV infection. The World Health Organization (WHO) ART guidelines focus on three classes of antiretroviral drugs, namely nucleoside or nucleotide reverse transcriptase inhibitors (NRTI), non-nucleoside reverse transcriptase inhibitors (NNRTI) and protease inhibitors. Two of the most common medications given as first-line treatment are the NNRTIs, efavirenz (EFV) and nevirapine (NVP). It is unclear which NNRTI is more efficacious for initial therapy. This systematic review was first published in 2010. OBJECTIVES: To determine which non-nucleoside reverse transcriptase inhibitor, either EFV or NVP, is more effective in suppressing viral load when given in combination with two nucleoside reverse transcriptase inhibitors as part of initial antiretroviral therapy for HIV infection in adults and children. SEARCH METHODS: We attempted to identify all relevant studies, regardless of language or publication status, in electronic databases and conference proceedings up to 12 August 2016. We searched MEDLINE, Embase, the Cochrane Central Register of Controlled Trials (CENTRAL), the World Health Organization (WHO) International Clinical Trials Registry Platform (ICTRP) and ClinicalTrials.gov to 12 August 2016. We searched LILACS (Latin American and Caribbean Health Sciences Literature) and the Web of Science from 1996 to 12 August 2016. We checked the National Library of Medicine (NLM) Gateway from 1996 to 2009, as it was no longer available after 2009. SELECTION CRITERIA: We included all randomized controlled trials (RCTs) that compared EFV to NVP in people with HIV without prior exposure to ART, irrespective of the dosage or NRTI's given in combination.The primary outcome of interest was virological success. Other primary outcomes included mortality, clinical progression to AIDS, severe adverse events, and discontinuation of therapy for any reason. Secondary outcomes were change in CD4 count, treatment failure, development of ART drug resistance, and prevention of sexual transmission of HIV. DATA COLLECTION AND ANALYSIS: Two review authors assessed each reference for inclusion using exclusion criteria that we had established a priori. Two review authors independently extracted data from each included trial using a standardized data extraction form. We analysed data on an intention-to-treat basis. We performed subgroup analyses for concurrent treatment for tuberculosis and dosage of NVP. We followed standard Cochrane methodological procedures. MAIN RESULTS: Twelve RCTs, which included 3278 participants, met our inclusion criteria. None of these trials included children. The length of follow-up time, study settings, and NRTI combination drugs varied greatly. In five included trials, participants were receiving concurrent treatment for tuberculosis.There was little or no difference between EFV and NVP in virological success (RR 1.04, 95% CI 0.99 to 1.09; 10 trials, 2438 participants; high quality evidence), probably little or no difference in mortality (RR 0.84, 95% CI 0.59 to 1.19; 8 trials, 2317 participants; moderate quality evidence) and progression to AIDS (RR 1.23, 95% CI 0.72 to 2.11; 5 trials, 2005 participants; moderate quality evidence). We are uncertain whether there is a difference in all severe adverse events (RR 0.91, 95% CI 0.71 to 1.18; 8 trials, 2329 participants; very low quality evidence). There is probably little or no difference in discontinuation rate (RR 0.93, 95% CI 0.69 to 1.25; 9 trials, 2384 participants; moderate quality evidence) and change in CD4 count (MD -3.03; 95% CI -17.41 to 11.35; 9 trials, 1829 participants; moderate quality evidence). There may be little or no difference in treatment failure (RR 0.97, 95% CI 0.76 to 1.24; 5 trials, 737 participants; low quality evidence). Development of drug resistance is probably slightly less in the EFV arms (RR 0.76, 95% CI 0.60 to 0.95; 4 trials, 988 participants; moderate quality evidence). No studies were found that looked at sexual transmission of HIV.When we examined the adverse events individually, EFV probably is associated with more people with impaired mental function (7 per 1000) compared to NVP (2 per 1000; RR 4.46, 95% CI 1.65 to 12.03; 6 trials, 2049 participants; moderate quality evidence) but fewer people with elevated transaminases (RR 0.52, 95% CI 0.35 to 0.78; 3 trials, 1299 participants; high quality evidence), fewer people with neutropenia (RR 0.48, 95% CI 0.28 to 0.82; 3 trials, 1799 participants; high quality evidence), and probably fewer people withrash (229 per 100 with NVP versus 133 per 1000 with EFV; RR 0.58, 95% CI 0.34 to 1.00; 7 trials, 2277 participants; moderate quality evidence). We found that there may be little or no difference in gastrointestinal adverse events (RR 0.76, 95% CI 0.48 to 1.21; 6 trials, 2049 participants; low quality evidence), pyrexia (RR 0.65, 95% CI 0.15 to 2.73; 3 trials, 1799 participants; low quality evidence), raised alkaline phosphatase (RR 0.65, 95% CI 0.17 to 2.50; 1 trial, 1007 participants; low quality evidence), raised amylase (RR 1.40, 95% CI 0.72 to 2.73; 2 trials, 1071 participants; low quality evidence) and raised triglycerides (RR 1.10, 95% CI 0.39 to 3.13; 2 trials, 1071 participants; low quality evidence). There was probably little or no difference in serum glutamic oxaloacetic transaminase (SGOT; MD 3.3, 95% CI -2.06 to 8.66; 1 trial, 135 participants; moderate quality evidence), serum glutamic- pyruvic transaminase (SGPT; MD 5.7, 95% CI -4.23 to 15.63; 1 trial, 135 participants; moderate quality evidence) and raised cholesterol (RR 6.03, 95% CI 0.75 to 48.78; 1 trial, 64 participants; moderate quality evidence).Our subgroup analyses revealed that NVP slightly increases mortality when given once daily (RR 0.34, 95% CI 0.13 to 0.90; 3 trials, 678 participants; high quality evidence). There were little or no differences in the primary outcomes for patients who were concurrently receiving treatment for tuberculosis. AUTHORS' CONCLUSIONS: Both drugs have similar benefits in initial treatment of HIV infection when combined with two NRTIs. The adverse events encountered affect different systems, with EFV more likely to cause central nervous system adverse events and NVP more likely to raise transaminases, cause neutropenia and rash.