Introduction: The use of HIV protease inhibitors (PIs) in a ritonavir (RTV)-boosted form is now common. However, randomized data comparing boosted with unboosted PI strategies are scarce. Methods: This randomized, open-label trial compared indinavir (IDV) 800mg three times daily with IDV/RTV 800/100 mg twice daily, both given with zidovudine (AZT)/lamivudine (3TC) twice daily in individuals with at least 3 months previous AZT experience. The primary endpoint was the time-weighted average change in HIV RNA from baseline. Designed as a 48-week study, follow-up continued until week 112. Primary analysis is by intention to treat.Results: One hundred and three patients commenced therapy and are included in the analysis. Patients had a median of 29 months past nucleoside reverse transcriptase inhibitor (NRTI) exposure. Baseline median (interquartile range) log10 HIV RNA was 4.0 (3.3–4.5) and CD4+ T-cell count 166 (40–323) cells/μl. After 112-weeks of study there was no significant difference observed between arms in the mean (SD) change in timeweighted average HIV RNA from baseline (–1.6 [1.1] HIV RNA copies/week/ml three times daily arm; –1.4 [1.1] HIV RNA copies/week/ml twice daily arm; P=0.3). Both arms were associated with substantial toxicity expressed as serious adverse events and study drug interruptions. The twice daily arm experienced greater dyslipidaemia. Mean (SD) changes in time-weighted CD4+ T-cell count from baseline were similar [88 (84) cells/week/μl three times daily arm; 70 [109] cells/week/μl twice daily arm; P=0.3).Conclusions: RTV-boosted IDV 800/100 mg twice daily demonstrated comparable efficacy to unboosted IDV 800mg three times daily dosing. Both regimens were associated with substantial toxicity. Use of lower doses of RTV-boosted IDV may result in better tolerability without loss of efficacy and warrant further research.