A prospective study was conducted to determine (i) the degree of yeast colonization in surgical patients with intra-abdominal perforations and (ii) whether the frequency of colonization is different in patients with a complicated postoperative course than in patients recovering uneventfully. A total of 1,496 specimens taken per- and post-operatively from the mouth, stomach, feces, urine, trachea, and abdomen of 109 surgical patients with intra-abdominal perforations were examined. Yeast was recovered from 98 (90%) of the patients and from 634 (42%) of the specimens. Approximately 70% of the specimens from the mouth and stomach, 47% of fecal specimens, and 31% of abdominal specimens were positive for yeast. A total of 42 patients had a complicated postoperative course. The majority of these patients were colonized with yeast at multiple body sites: yeast was recovered on one or more occasions from two or more body sites in 90% and from three or more body sites in 71%. Many of the patients with an uncomplicated postoperative course also were colonized: yeast was recovered from two or more body sites in 69% and from three or more body sites in 34%. The results of this study indicate that treatment recommendations based on yeast colonization will expose a large number of patients to unnecessary or even harmful antifungal treatment. This does not mean that yeast colonization is insignificant; however, more accurate criteria and methods based on prospective clinical studies are needed to detect patients at risk of developing severe Candida infection.