OBJECTIVES: In Brazil, switching to a protease inhibitor (PI) based ARV regimen is recommended as second line therapy for experienced patients failing non-nucleoside reverse transcriptase inhibitors. The BMS-045 study compared ATV+RTV and LPV/r regimens in ARV-experienced patients.  Similar viral load (VL) suppression rates <400 copies were reported, but LPV/r provided greater suppression rate <50 copies. Total cholesterol (TC) levels improved to guideline levels in 23% of ATV+RTV patients and became elevated in 7% of LPV/r patients at 48 weeks. The long term clinical and cost impact of this difference is not yet clear.  The objective of this study was to examine the long term HRQL and economic implications in Brazil for LPV/r versus ATV+RTV treatment of ARV-experienced patients. METHODS: A previously published HIV Markov model was adapted.  Baseline assumptions: TC profile and CD4 cell distribution matching the BMS-045 population. HRQL and survival outcomes were measured in quality adjusted life years (QALYs). Costs in Brazilian Reale were indexed to 2011.  ARV costs and HIV treatment patterns were based on Brazilian references.  Lifetime costs/outcomes were discounted at 3% per annum. A national health services perspective was adopted. RESULTS: VL suppression differences favored LPV/r, driving a net improvement in survival (0.31 QALYs, 106 days).  Five and 10 year cost savings (BRL1,816, BRL1,496 per patient) were projected for LPV/r. Lifetime costs were slightly higher for LPV/r due to improved survival. An incremental cost effectiveness ratio (ICER) of BRL2319 per QALY gained was estimated for the LPV/r regimen, which is highly acceptable by Brazilian threshold . CONCLUSIONS: Compared to ATV+RTV, an LPV/r based regimen is cost saving through the first 10 years of survival and is a cost effective use of public resources for ARV-experienced Brazilian patients. LPV/r implementation is supported by its improved viral suppression, short/long term cost savings and favorable ICER.