Objectives: The increased cardiovascular risk historically associated with protease inhibitor (PI) use in HIVinfected individuals is only partly attributable to fasting dyslipidaemia, and may not apply to the currently recommended PI drugs-atazanavir and darunavir. In this prospective pilot study, we aimed to describe the isolated metabolic differences between atazanavir/ritonavir and darunavir/ritonavir on circulating fasting and post-prandial lipids (primary objective), and arterial stiffness, glycaemic and inflammatory markers (secondary objectives) by administering them as monotherapy to healthy, HIV-uninfected adults. Methods: Participants were randomised 1:1 to receive open-label atazanavir (300 mg/day) or darunavir (800 mg/day), both boosted with ritonavir (100 mg/day), for 4 weeks. A standardised high-energy meal was consumed at weeks 0 and 4, with hourly post-prandial blood samples and arterial stiffness assessed for 4 hours. Mean fasting and post-prandial measurements were compared, the latter as the change in mean post-prandial incremental area under the curve. Results: Twenty participants (10 atazanavir/ritonavir, 10 darunavir/ritonavir; 50% male, mean age 38 years, mean blood pressure 113/72 mmHg, mean total cholesterol 4.4 mmol/L, mean triglycerides 0.9 mmol/L) completed the study. After four weeks, fasting total cholesterol rose more with darunavir/ritonavir than atazanavir/ritonavir (+0.8 mmol/L vs. +0.3 mmol/L, respectively; p=0.041), as did low-density lipoprotein cholesterol (+0.1 mmol/L vs. +0.7 mmol/L, respectively; p=0.017). Between-group differences in post-prandial lipid changes were non-significant. Postprandial decline in arterial stiffness was greater with atazanavir/ritonavir than darunavir/ritonavir (-27.6 h% vs. +0.1 h%, respectively; p=0.041). Bilirubin rose significantly with atazanavir/ritonavir but not darunavir/ritonavir (+29 µmol/L vs.-0.6 µmol/L, respectively; p=0.001). Changes in other parameters were similar. Conclusion: Atazanavir/ritonavir has modestly, but significant, favourable metabolic and haemodynamic profiles over darunavir/ritonavir, evident even after short-term monotherapy. Post-prandial lipid effects were similar, however. Further study of treatment-emergent cardiovascular event incidence with darunavir/ritonavir therapy is required to determine the clinical relevance of this between-group difference. J o ur nal o f A ID S & Cli n ic a l R es earc h