OBJECTIVE: Infections caused by Candida spp. are a major cause of morbidity and mortality in critically ill patients and usually develop from endogenous colonization. We assessed the effectiveness of adding fluconazole to a selective digestive decontamination regimen to prevent candidal infections. DESIGN AND SETTING: We performed a prospective, randomized, double-blind, placebo-controlled trial among medical and surgical intensive care unit patients at a large university hospital. PATIENTS: All adult patients mechanically ventilated for at least 48 h with an expectation to remain so for at least an additional 72 h, and receiving selective decontamination of the digestive tract. INTERVENTIONS: Patients were randomly assigned fluconazole 100 mg daily (n=103) or placebo (n=101). MEASUREMENTS AND RESULTS: Candida infections occurred less frequently in the fluconazole group (5.8%) than in the placebo group (16%; rate ratio 0.35; Cl(95) 0.11-0.94). Some 90% of candidemia episodes occurred in the placebo group (rate ratio for fluconazole use 0.10; Cl(95) 0.02-0.74). The rate of treatment failure, development of candidal infection, or increased colonization, was 32% in the fluconazole group and 67% in the placebo group (P<0.001). Crude in-hospital mortality was similar in the two groups (39% fluconazole vs. 41% placebo). CONCLUSIONS: Prophylactic use of fluconazole in a selected group of mechanically ventilated patients at high risk for infection reduces the incidence of Candida infections, in particular candidemia.