Background: Although the increased frequency of invasive fungal infections, there are limited data on factors that influence the use of antifungals. The aim of this study was determinate the factors associated with use of antifungal therapy in patients in the intensive care units from third level hospitals in Colombia. Methods: An observational, analytic, nested, case-control study was made, based on hospitalized patients. We made a prospective revision of medical records of adult patients in the intensive care unit from 5 tertiary care hospitals in Colombia during 3 months. Epidemiological, clinical and microbiological variables were followed. The case group were conformed by patients with antifungal therapy, while control group were formed by patients without antifungal therapy, paired by date of admission and time stay (3 controls per each case). A multivariable logistic regression analysis was performed with antifungal use as dependent variable. Results: 492 patients were followed (median hospitalization time 8 days), 53,3% (n = 263) were male. 26,6% had an infectious diagnosis at admission, 19,3% (n = 97) diabetes, 12% (n = 59) cancer, and 8,7% (n = 43) immunosuppression. Mean Apache score at admission was 12,3 (SD = 6,7), candidemia was detected in 1.3% of the patients (n = 6), andCandidacolonization in 2,2% (n = 11). 26 patients received antimycotic treatment (5.2%), with an incidence of 61.1 days of use X 1000 patient-days. Factors associated with antifungal use in the multivariated model were: at least one site of colonization (OR = 54,7 CI 95% = 2,4 – 1273), medical order for fungal culture (OR = 19,6 CI 95% = 2,5 - 151.2), use of inotropic medications(OR = 39,4 CI 95% = 4,6 – 334,5), parenteral nutrition (OR = 50,4 IC95) and fever over 48 hours (OR = 78,8 CI 95% = 3,8 – 1644). Conclusion: Factors associated with antifungal use were related to the severity of the clinical condition of the patient. Risk factors for fungal infections did not usually determined antifungal use.