OBJECTIVES: Fatigue is a common, distressing symptom in HIV+ patients. This analysis evaluates whether substitution to LPV/r, a generally well tolerated and efficacious PI, affects fatigue in HIV+ patients experiencing Grade 2 side effects (SE) attributed to their PI/NNRTI. METHODS: In the open-label PLATO trial, patients experiencing Grade 2 PI/NNRTI-associated SE were randomized (4:1) to immediate substitution (IS) of their PI/NNRTI with LPV/r at baseline or deferred substitution (DS) at Week 4 (Wk4). The MOS-HIV and ACTG Symptoms Distress Module, with 2 additional items for nephrolithiasis (ASDM), were administered at baseline, Wk4 and Wk8. Fatigue was measured by MOS-HIV fatigue-domain and ASDM fatigue/bothersomeness-item. Sleep-disorder was measured by ASDM sleep-disorder-item. The Center for Epidemiologic Studies-Depression (CES-D) questionnaire was administered at baseline and Wk8. RESULTS: Eight hundred twenty-seven patients previously on nelfinavir (n = 291), indinavir (n = 170), indinavir/ritonavir (n = 182), efavirenz (n = 136) or another PI/NNRTI (n = 48) were analyzed (80% male, mean age 42 yrs, 75% with baseline HIV RNA < 50 copies/mL). At baseline, mean MOS-HIV fatigue-domain score was 56.7, with 62.3% rating fatigue as bothersome. Baseline fatigue scores were correlated (p <0.05) with presence of depression (CES-D> = 16), sleep-disorder, and years since HIV diagnosis. At Wk4, improved fatigue scores were seen in IS vs. DS groups (MOS-HIV fatigue-domain: + 8.711 vs. + 0.068, p <0.001; ASDM fatigue/bothersomeness-item: -0.486 vs. + 0.074, p <0.001), irrespective of prior PI/NNRTI regimens. At Wk8, fatigue improvement remained for IS group, while DS group began to improve. Improved fatigue scores were associated (p <0.05) with IS, reduced prevalence of depression, and improved sleep-disorder scores. Improvement in fatigue and IS were significant predictors of improved MOS-HIV physical health summary score at Wk4 (p <0.05). CONCLUSIONS: Fatigue scores were improved following substitution with LPV/r, and were associated with reduced prevalence of depression and improved sleep-disorder scores. Improvement in fatigue was independent of prior PI/NNRTI and was a predictor of improved physical health.