Background: Candida peritonitis (CP) is generally considered to be a severe disease, but its impact on outcome in critically ill patients remains unknown. Hypothesis: The predictive factors of mortality due to CP can be determined by study of a, population of patients with CP. Design: A retrospective review of a prospective surgical intensive care unit (ICU) database of patients (January 1, 1994, through December 31, 2000). Setting: University hospital in Paris, France. Patients: Eighty-three patients with generalized CP. Main Outcome Measures: Demographic and microbiologic data and outcome were collected, and nonsurvivors were compared with survivors. Results: Overall ICU mortality due to CP was 43 (52%) of 83 patients. In a stepwise multivariate logistic regression, the following 4 variables were independently associated with mortality: APACHE II (Acute Physiology and Chronic Health Evaluation 11) score on admission of at least 17 (odds ratio [OR], 28.4; 95% confidence interval [CI], 5.7-142.5; P < .001), respiratory failure on admission (OR, 10.6; 95% CI, 2.2-51.2; P = .003), upper gastrointestinal tract site of peritonitis (OR, 7.7; 95% CI, 1.7-34.7; P = .007), and results of direct examination of peritoneal fluid that were positive for Candida (OR, 4.7; 95% CI, 1.2-19.7; P = .0.02). Conclusions: These results confirm the severity of CP in ICU patients and emphasize the prognostic value of direct examination of peritoneal fluid for Candida in this context.