The long-term efficacy of once-daily (qd) fosamprenavir (FPV) 1400 mg boosted by ritonavir 100 mg (FPV/r100) has not been evaluated previously. A 96-week open-label, randomized, multicenter study compared the efficacy/safety of FPV/r100 with FPV 1400 mg boosted by ritonavir 200 mg qd (FPV/r200), plus abacavir/lamivudine 600 mg/300 mg qd, in antiretroviral-naive, HIV-infected patients with viral load (VL)> or =1000 copies/ml. Primary endpoints were proportion of patients achieving VL <400 copies/ml or discontinuing for drug-related reasons. In the intent-to-treat:exposed (ITT-E) population, missing = failure (M = F), and observed approaches were used to assess between-arm differences in VL responses by Cochran-Mantel-Haenszel test and CD4(+) count by Wilcoxon rank-sum test. One hundred and fifteen (115) patients enrolled, with 58 on FPV/r100 (median VL 4.7 log(10) copies/ml; CD4(+) count 259 cells/mm(3)) and 57 on FPV/r200 (median VL 4.9 log(10) copies/ml; CD4(+) count 179 cells/mm(3)). Fewer FPV/r100-treated patients discontinued treatment prematurely (12 vs. 24) and experienced virologic failure (5 vs. 8, none developing major protease inhibitor resistance mutations). At week 96, more FPV/r100-treated patients had VL <400 copies/ml [ITT-E,M = F: 78% (45/58) vs. 53% (30/57), p = 0.006; observed: 98% (45/46) vs. 94% (30/32)] and VL<50 copies/ml [ITT-E,M = F: 66% (38/58) vs. 53% (30/57); observed: 83% (38/46) vs. 94% (30/32)]. The FPV/r100 and FPV/r200 arms were similar at week 96 regarding median change from baseline in CD4(+) count (+265 vs. +260 cells/mm(3)) and total cholesterol (+33 vs. +35 mg/dl), and in total-cholesterol:HDL-cholesterol ratio (4.0 vs. 4.1) and type/frequency of treatment-related grade 2-4 adverse events, although FPV/r100 was associated with a lower elevation in triglycerides (+27 vs. +48 mg/dl). In conclusion, through 96 weeks, FPV/r100 was more effective and prompted less elevation in triglycerides than FPV/r200.