Objectives: Virological failure on first-line nonnucleoside reverse transcriptase inhibitor (NNRTI)-based treatment regimens has become a problem in HIV-infected children on long-term antiretroviral therapy (ART). Protease inhibitor (PI)-based regimens are therefore often given to children failing NNRTI-based regimens. The aim of the study was to assess the 48-week effectiveness, safety and predictive factors for viral suppression of PI-based regimens in HIV-infected Thai children who had failed NNRTI-based regimens. Methods: This study assessed 41 HIV-infected children who had failed first-line NNRTI-based regimens and were switched to PI-based regimens for at least 48 weeks. We assessed their CD4 cell counts, plasma HIV RNA levels, weight-for-age and height-for-age z-scores, and adverse events. Results: The children's median age was 9.5 years (range 1.5-15.8 years). At baseline, their median CD4 cell count was 276 cells/μL [interquartile range (IQR) 160-749 cells/μL], and their median plasma HIV RNA level was 4.5 log10 HIV-1 RNA copies/mL (IQR 3.9-4.8 log10 copies/mL). After 48 weeks of PI-based therapy, their CD4 cell counts increased to a median of 572 cells/μL (IQR 343-845 cells/μL) and in 73.2% plasma HIV RNA levels decreased to <50 copies/mL. Their median weight-for-age and height-for-age z-scores were stable over the period of the study. Diarrhoea occurred in 29.3% of patients. Triglyceride levels were significantly higher at weeks 24 and 48 in comparison to baseline measurements. Conclusions: PI-based regimens are safe and effective for HIV-infected Thai children who have failed first-line NNRTI-based regimens. However, long-term follow-up is warranted in order to ascertain the feasibility and sustainability of these new regimens. © 2012 British HIV Association.