Increased lopinavir (LPV) exposure obtained in vivo through combination with low-dose ritonavir may overcome a certain grade of resistance but not all. We sought to analyze LPV variability and possible risk factors. LPV trough plasma concentrations were determined by high-performance liquid chromatography after 1, 4, and 12 weeks from salvage regimens and tested in both univariate and multivariate regression analyses with age, gender, weight, risk factors for HIV acquisition, hepatitis C virus reactivity, hepatitis B surface antigen positivity, baseline aspartate transferase (AST) or alanine transferase (ALT) levels, creatinine, non-nucleoside reverse transcriptase inhibitors (NNRTIs) or tenofovir as concomitant drugs, and NNRTIs administered in the previous regimen. Fifty-six patients were included into the study. Among them, 8 of 56 (14.3%) at week 1, 12 of 56 (21.4%) at week 4, and 9 of 56 (16.1%) at week 12 had suboptimal LPV plasma concentrations, defined as trough concentration less than 4 microg/mL. No correlation was found between LPV trough concentrations and assessed variables. In conclusion, pharmacokinetic variability and low LPV concentrations have been found, supporting the use of therapeutic drug monitoring in those starting this drug.