OBJECTIVE To assess the tolerability of liposomal amphotericin B (L-AmB) in critically ill patients with elevated serum creatinine concentrations (Cr) (> 1.5 mg/dL) at starting L-AmB therapy. METHODS Retrospective, multicenter, comparative study of two cohorts of critically ill patients treated with L-AmB during 3 or more days, the difference between them was the level of Cr at the beginning of treatment. A cutoff value of Cr of 1.5 mg/dL was established. Patients undergoing extrarenal depuration procedures before or 48 hours after starting L-AmB were excluded. The primary endpoint was the difference between Cr values at the end of treatment as compared with Cr at starting L-AmB. Secondary endpoints were treatment-related withdrawals, need of extrarenal depuration techniques, and treatment-related severe adverse events. Demographic data, underlying illness, indication of L-AmB therapy, concomitant risk factors of nephrotoxicity, and vital status at ICU and hospital discharge were recorded. RESULTS A total of 122 patients admitted to 26 ICUs (16 with Cr > 1.5 g/dL; 106 with normal Cr levels) were recruited. Main reasons for the use of L-AmB in both groups were the broad spectrum of the drug and the presence of hemodynamic instability. L-AmB was administered as first-line treatment in 68.8% of patients with elevated Cr and in 52.8% with normal Cr. The APACHE II score on ICU admission was 25 in patients with elevated Cr and 17 in those with normal Cr values (p < 0.001). Duration of treatment with L-AmB was 16 and 12 days in patients with elevate and normal Cr values, respectively, with a mean dose of 3.5 vs 3.9 mg/kg/day. The use of concomitant nephrotoxic drugs, mortality rate, and ICU and hospital length of stay were similar in both cohorts. In patients with renal function impairment at the initiation of L-AmB treatment, an absolute decrease of Cf-Ci of 1.08 mg/dL was observed (P < 0.001). A decrease of Cr levels to normal limits was observed in 50% of the patients; in 37.5% of patients there was a decrease but normal levels were not achieved, whereas a Cr increased occurred in only one (6.25%) patient. None of the patients required withdrawal of L-AmB or use of extrarenal depuration procedures. Treatment-related severe adverse events were not reported. CONCLUSIONS In critically ill patients with impaired renal function, the impact of L-AmB on renal function was minimal. L-AmB can be used for the treatment of fungal infections in critically ill patients independently of renal function at the initiation of treatment.