OBJECTIVE: Clinical trials results are used to guide empiric antifungal therapy. However, how subjects enrolled into clinical trials compare to “real-world” patients is unclear. The purpose of this study was to compare the inclusion/exclusion criteria from published candidemia trials to hospitalized patients with candidemia at our institution. METHODS: The inclusion/exclusion criteria from two large, randomized trials of candidemia (N Eng J Med 1994;331:1325-30 and Clin Infect Dis 2003;36:1221-8) were assessed for 86 patients hospitalized with candidemia between 2002 to 2004 at our institution. Variables collected included APACHEII, vitals, receipt of antifungals, past medical and hospitalization history including hematologic cancer, HIV, burns, allergy, organ transplantation, previous candidemia and laboratory values including bilirubin and liver function tests. Inclusion/exclusion criteria from the published clinical trial were compared to results from our patient population. Percent of patients that would be eligible for study entry were calculated. RESULTS: Seventy-nine percent of our patients would have been excluded from the clinical trials. The most common exclusion criteria included receipt of amphotericin before identification of candida (42%), laboratory abnormalities including elevated bilirubin (15%) and elevated aminotransferases (2%), organ transplantation (8%) or HIV (3%). No patients were excluded due to pregnancy, burns, allergies, other investigational agents, or congenital immunodeficiency syndrome. If it were possible to include patients before receipt of antifungals, 51% of patient would have been able to be enrolled in the study. ApacheII scores were similar in published clinical trials (16.1) compared to our patient population (17.3). CONCLUSION: Subjects enrolled into candidemia clinical trials were similar to our patient population including severity of illness. Exclusion criteria generally involve laboratory safety parameters or different patient populations. Results from these clinical trials are applicable to "real-world" patients.