Background.The efficacy and tolerance of switching from zidovudine (ZDV) and lamivudine (3TC) in clinically stable HIV-infected children with incomplete viral suppression to stavudine (d4T), nevirapine (NVP) and ritonavir (RTV) has not been determined. Aim.To evaluate the safety, tolerance, antiviral activity and immunologic changes after the change to a three drug combination.Methods.During a clinical trial in which HIV-infected antiretroviral-experienced children were initially randomized to receive d4T/RTV, ZDV/3TC/RTV or ZDV/3TC (Step 1), 48 children who had HIV RNA ≥10 000 copies/ml after ≥12 weeks of ZDV/3TC therapy in Step 1 were switched to d4T/NVP/RTV in Step 2. The proportion of children receiving therapy with HIV RNA ≤400 copies/ml at Study Weeks 24 and 48 receiving d4T/NVP/RTV in Step 2 were compared with children receiving RTV-containing regimens in Step 1.Results.At 24 weeks of treatment with d4T/NVP/RTV in Step 2, 48% (23 of 48) of children had HIV RNA ≤400 copies/ml compared with 34% (31 of 92) and 47% (44 of 93) receiving d4T/RTV or ZDV/3TC/RTV for 24 weeks in Step 1; at 48 weeks virologic response was 44, 27 and 42% in Step 2 d4T/NVP/RTV, Step 1 d4T/RTV and Step 1 ZDV/3TC/RTV arms, respectively.Conclusions.A delay of 7 to 12 months in the initiation of protease inhibitor-containing combination therapy in children receiving dual nucleoside analogue therapy did not adversely affect the RNA response during the first 48 weeks of treatment.