Objectives: To assess the most frequent resistance-associated mutations (RAMs) to lopinavir/ritonavir in a cohort of patients attended in daily practice. Methods:We retrospectively identified 195 multitreated subjects with virological failure. Patients were classified as follows: (i) 71 (36.4%) never received lopinavir/ritonavir (lopinavir/ritonavir naive); (ii) 75 (38.5%) had previously failed on lopinavir/ritonavir; and (iii) 49 (25.1%) were on lopinavir/ritonavir at failure. RAM patterns were assessed. Medians, IQRs, percentages, Kruskal-Wallis, χ 2 or Fisher's exact test, and multinomial logistic regression were used whenever appropriate. Results: L10I/F, K20R, L24I, L33F, M36I, M46I/L, I47V, G48V, F53L, I54V, A71V, G73S, V82A, I84V and L90M (all with P≤0.037) were protease RAMs overexpressed in patients with lopinavir/ritonavir failure. L10I, M36I, M46I, I54V, L63P, A71V, V82A, I84V and L90M were the most common in lopinavir/ritonavir-naive patients. Other IASUSA RAMs for lopinavir/ritonavir (L10R/V, K20M, V32I, I47A, I50V, I54L/A/M/T/S, A71T, L76V and V82F/T/S) were not associated with previous or current failure to lopinavir/ritonavir. Lopinavir/ritonavir failure was associated with the number of protease RAMs (OR=1.146, 95% CI=1.287, 1.626), higher exposure to protease inhibitors, and the presence of E44D, L33F, I54V and I84V. Conclusions: In multitreated patients with previous or current lopinavir/ritonavir failure, some protease mutations are selected at significantly greater rates. L10I, M36I, I54V, L63P, A71V, V82A and L90M were found in >50% of cases. Thus, their presence should be expected when genotypic testing results are not available. The number of protease RAMs and higher prior exposures to protease inhibitors were significantly associated with lopinavir/ritonavir failure.