Abstract Objective To analyze the effect of switching the ritonavir-boosted protease inhibitor (PI/r) in a stable combined antiretroviral therapy (cART) regimen to raltegravir on low-density lipoprotein (LDL) particles, and lipoprotein-associated phospholipase A2 (Lp-PLA2). Design Substudy of a multicenter randomized trial that compared the efficacy of switching a PI/r to raltegravir-based cART in stable HIV-infected patients. Methods LDL size and phenotype (by gel-gradient electrophoresis), Lp-PLA2 (by 2-thio-PAF Cayman), proprotein convertase subtilisin/kexin type 9 (PCSK9) (by ELISA), and standard lipid parameters were measured at baseline and week 48. Results Eighty-one (PI/r n  = 41 and raltegravir n  = 40) patients were evaluated. No differences in baseline demographic and metabolic variables between arms were found except in apolipoprotein (Apo) B ( p  = 0.042). At week 48, total cholesterol (TC) ( p  < 0.001), LDL-c ( p  = 0.023), non-high density lipoprotein cholesterol non-high-density lipoprotein cholesterol (non-HDL-c) ( p  < 0.001), TC/HDL ( p  = 0.026), triglyceride ( p  < 0.001), Apo B ( p  < 0.001), Apo A-I ( p  = 0.004) and Lp (a) ( p  = 0.005) decreased in raltegravir arm compared to PI/r arm. At week 48, a shift from LDL phenotype B to the less atherogenic phenotype A was observed only in raltegravir arm ( p  < 0.001). LDL size increased (PI/r 2.1 nm, p  = 0.019; raltegravir 3.8 nm, p  = 0.001) and cholesterol content in small and dense LDL subfractions (LDL 4,5,6) decreased (PI/r p  = 0.007, raltegravir p  = 0.006) at week 48 in both arms. Total Lp-PLA2 activity (PI/r p  = 0.037 and raltegravir p  = 0.051) and PCSK9 plasma concentration decreased in both arms (PI/r p  = 0.034 and raltegravir p  < 0.001). Conclusions Switching a PI/r to a raltegravir-based cART in virologically suppressed HIV-infected patients was associated with an overall improvement in lipid profile, including a shift to a less atherogenic LDL phenotype.