Maintenance with a triple nucleoside reverse transcriptase Inhibitor (NRTI) regimen after successful induction with a dual NRTI/protease inhibitor (PI) combination may be advantageous, because of low pill burden, favorable lipids, and less drug interactions. This strategy to become free of PI-related problems without losing viral efficacy has not been formally tested. We performed a randomized, open-label, multicenter, 96-week comparative study in antiretroviral therapy (ART)-naïve patients with CD4 <or=350 cells/mm(3) and HIV-1 RNA concentrations (viral load [VL]) greater than 30,000 copies per milliliter. Patients were randomized after reaching VL less than 50 copies per milliliter on two consecutive occasions between 12 and 24 weeks after start of zidovudine/lamuvidine and lopinavir/ritonavir combination. Eligible subjects switched to abacavir/lamivudine/zidovudine (TZV) or continued the PI-containing regimen. Here we present the 48-week data with virologic success rate (failure: VL > 50 copies per milliliter). Two hundred seven patients had similar baseline (BL) characteristics: median CD4 180 cells/mm(3), median VL 5.19 log(10) copies per milliliter. One hundred twenty subjects (58%) met randomization criteria. Baseline VL differed significantly between dropouts and randomized subjects (median 5.41 versus 5.06 log(10) copies per milliliter, p = 0.017), as did CD4 cells (median 160 and 200 cells/mm(3), p = 0.044). Sixty-one subjects received TZV and 59 subjects continued NRTIs/PI. At week 48, 2 patients in the TZV group and 5 in the PI group did not have a sustained virologic suppression (log rank test; p = 0.379). CD4 counts increased significantly in both arms. In ART-naïve patients, TZV maintenance had similar antiviral efficacy compared to continued standard ART at 48 weeks after baseline. Patients on successful standard ART can be safely switched to a NRTI-only regimen, at least for the tested time period.