The human immunodeficiency virus (HIV) has become one of the greatest challenges to global public health. In 2007 UNAIDS estimated that 33.2 million people were living with HIV. Currently recommended regimens for initiating HIV treatment consist of either a non-nucleoside reverse transcriptase inhibitor (NNRTI) or ritonvair-boosted protease inhibitor (PI) combined with two nucleoside reverse transcriptase inhibitors (NRTIs); however, there may be some patients for whom NNRTIs and PIs may not be appropriate. The aim of this review was to evaluate the effects of Trizivir, a fixed-dose combination of three NRTIs (abacavir-lamivudine-zidovudine) for initial treatment of HIV infection. In February 2008, we searched the Cochrane Library, PubMed, EMBASE, AIDSearch and GATEWAY and checked reference lists of identified articles. In May 2009, we repeated the search in PubMed and the Cochrane Library. We selected randomized controlled trials (RCTs) with a minimum follow-up time of six months which compared Trizivir with either a PI- or NNRTI-based therapy among antiretroviral-naive HIV-infected patients aged at least 13 years. Three authors independently extracted data. We calculated the relative risk (RR) or mean difference (as appropriate) for each outcome with its 95% confidence interval (CI) and conducted meta-analysis using the random-effects method because of significant statistical heterogeneity (P<0.1). We identified nine potentially eligible RCTs, three of which met our inclusion criteria. One trial compared Trizivir to efavirenz (an NNRTI) plus two or three NRTIs; the second trial compared Trizivir to a treatment based on the PI nelfinavir; and the third compared Trizivir to atazanavir (a PI) plus two NRTIs. Overall, there was no significant difference in the incidence of virological failure between participants on Trizivir and those on PI-based or NNRTI-based therapy (three trials, N=1687; RR 1.14, 95% CI 0.56 to 2.32). However, there was significant heterogeneity between the results of the three trials (heterogeneity P=0.009, I(2)=79%), with a significant increase in virological failure for Trizivir compared to efavirenz (N=1147; RR 1.93, 95% CI 1.46 to 2.55) but no difference between Trizivir and PIs (two trials, N=540; RR 0.82, 95% CI 0.50 to 1.36). We found no significant differences between Trizivir and either the PI or NNRTI on CD4+ cell counts (standardized mean difference -0.01, 95% CI -0.11 to 0.09, heterogeneity P=0.59, I(2)=0%), severe adverse events (RR 1.41, 95% CI 0.61 to 3.25, heterogeneity P=0.03, I(2)=73%) and hypersensitivity reactions (RR 4.04, 95% CI 0.41 to 40.02, heterogeneity P=0.03, I(2)=72%). Only the studies involving PIs reported the effect of the treatment regimens on the lipid profile. One study found that at 96 weeks, the mean increase in total cholesterol from baseline was significantly lower with Trizivir than with nelfinavir, but there were no significant differences with triglyceride levels. The second study found the fasting lipid profile to be comparable in both the Trizivir and atazanavir arms at 48 weeks. Our findings indicate that Trizivir remains a viable option for initiating antiretroviral therapy, especially in HIV-infected patients with pre-existing hyperlipidaemia and those who do not tolerate ritonavir.