OBJECTIVES: HIV guidelines consider LPV/r a preferred protease inhibitor for use during pregnancy. The budget implications of proactively initiating LPV/r versus initiating DRV+RTV and then potentially switching to LPV/r upon pregnancy were examined. METHODS: A cost minimization analysis was performed (US health care perspective) for HIV-1 infected, treatment-naïve WOCBA comparing: initiating LPV/r versus initiating DRV+RTV and switching to LPV/r when pregnant.  A discrete event simulation was employed to represent antiretroviral (ARV) therapy management. Clinical trial data were used to model pregnancy rates, ARV switch rates, treatment impact as a function of CD4-cell count/viral load, adherence, treatment response, acquired resistance mutations, and treatment changes. Five- and 10-year costs incurred due to ARV therapy, clinical management of AIDS-related, non-AIDS related, and cardiovascular events were estimated. Analysis assumptions:  switching to LPV/r can occur only once at first pregnancy, women’s medication adherence improves 15% at pregnancy and to 100% if viral rebound. Analyses varied the rate of switching to LPV/r at time of pregnancy (0%, 30%, 100%), pregnancy rates, adherence improvement, and health care costs. Daily drug cost (WAC): LPV/r + TDF/FTC, $56.59; DRV+RTV+TDF/FTC, $73.89. Costs were discounted 3% per annum. RESULTS: With 0% switch, survival was similar for LPV/r and DRV+RTV, 7.68 and 7.69 life years, respectively  (+/- 0.03 QALYs) at 10 years.  Five- and 10-year health care costs of ARV-naïve WOCBA who initiate LPV/r were $107,790 and $192,352 per patient, respectively, versus $132,694 and $235,854 when initiating DRV+RTV (a $43,502 per patient savings at 10 years). If 100% of patients who initiated DRV+RTV switched to LPV/r upon pregnancy, savings per patient were reduced 21.3%.  Sensitivity analyses showed that initiating LPV/r was always cost-saving relative to DRV+RTV. CONCLUSIONS: Initiating HIV infected, treatment-naïve WOCBA on LPV/r was cost saving compared to initiating DRV+RTV.  Analysis limitations include the uncertainty of long-term outcomes projections driven by short-term clinical trial endpoints.