A prospective, randomized study was conducted to determine if prophylactic antifungal agents prevented yeast colonization (YC) or yeast sepsis (YS), or if they diminished mortality in 292 critically ill adult (nontransplant/nonburned) surgical and trauma patients admitted to the SICU for 48 hours or longer. Patients were randomized to receive (group I) no therapy, (group II) clotrimazole 10 mg three times a day, (group III) ketoconazole 200 mg per day, or (group IV) nystatin 2 million units every 6 hours. For comparison patients were stratified by the criteria of Slotman and Burchard into high risk (3 risk factors) and low risk (<3 risk factors). Fifty patients (17%) had yeast colonization, nine (3.1%) had yeast sepsis, and 41 (14%) died. Stepwise logistic regression analysis of yeast colonization and sepsis using the variables APACHE II scores>10, need for ventilator support >48 hours, and 14 risk factors (Slotman and Burchard) showed that treatment with three or more antibiotics, APACHE ll>10, and ventilatory support >48 hours were the only three variables that were significant predictors of yeast colonization and sepsis. There was no significant difference between the four groups with regard to YC (23%, 18%, 12%, and 15%, respectively), YS (3%, 1%, 2%, and 7%, respectively), or mortality (15%, 14%, 6%, and 20%, respectively). In the 165 high-risk (3 risk factors) patients, although no difference was seen between the groups, a lower incidence of yeast colonization (21% vs. 41%, p<0.05) but not yeast sepsis (5.5% vs. 5.1%) was observed when all the prophylactically treated patients (n=126) were compared with the controls (n=39). Prophylactic use of these agents in only the high-risk (3 risk factors) nontransplant/ nonbumed trauma/surgical patients is possibly effective and needs to be evaluated in a large multicenter study.