OBJECTIVES: This study estimates lipid profile changes and abnormalities in treatment naïve HIV patients initiating atazanavir/ritonavir (ATV/r) versus darunavir/ritonavir (DRV/r) using a matched indirect comparison.  METHODS: Two similarly designed randomized trials were identified. Individual patient-level data were available for the CASTLE trial comparing ATV/r (n=430) and lopinavir/r (LPV/r) (n=438); published summary data were used from the ARTEMIS trial comparing DRV/r (n=343) and LPV/r (n=346). To adjust for cross-trial differences, CASTLE patients were re-weighted to match summary baseline characteristics in ARTEMIS. Lipid profile changes and serious (grade 2-4) lipid laboratory abnormalities (LA) at 48 weeks were compared between balanced trial populations after matching. If significant differences in LPV/r outcomes existed after matching, the differences between active treatments and LPV/r were compared across trials.  RESULTS: Data from all patients in the two trials were included. Before matching, CASTLE patients at baseline had a higher proportion of HIV-1 RNA>100,000 copies/mL, a lower proportion CDC class C, and lower median CD4 cell count than ARTEMIS patients. An unadjusted cross-trial comparison at 48 weeks showed that ATV/r and DRV/r had similar changes in total cholesterol, low density lipoprotein (LDL), and triglycerides, but ATV/r had significantly lower rates of total cholesterol LA (7% vs.13%, p=0.008) and LDL LA (8% vs. 13%, p=0.040). After balancing mean baseline characteristics, ATV/r and DRV/r at 48 weeks had similar changes in total cholesterol, LDL, and triglycerides (adjusted difference [AD]=  -7.19,  -7.11 and 7.59  mg/dL, respectively) and similar percentages of grade 2-4 triglycerides and cholesterol LAs (AD=0% and 1%, respectively); however, ATV/r had a significantly lower rate of serious LDL LA (AD=-9%, p=0.006). CONCLUSIONS: While the overall lipid profiles of ATV/r and DRV/r were similar at 48 weeks in this matching-adjusted indirect comparison, ATV/r had a significantly lower rate of serious LDL LA.