BACKGROUND: Recently, we noticed an increased isolation rate for Candida tropicalis from urine of the patients in the intensive care unit (ICU) of the neurosurgery department in our hospital. We inves-tigated the risk factors for funguria and performed randomly amplified polymorphic DNA (RAPD) analysis for the isolates. METHODS: A total of 27 strains including 12 strains of C. tropicalis from the urine of ICU patients collected during a one-month period, 13 strains from other wards than ICU, and 2 control strains were analyzed. RAPD analysis was performed using 2 primers (UBC78 and CD16S). Medical re-cords of the patients were reviewed to determine the risk factors for funguria by C. tropicalis. RESULTS: RAPD analysis showed an identical pattern for all strains of C. tropicalis from ICU patients and a heterogeneous pattern for the isolates from other wards. C. tropicalis funguria of these ICU patients was significantly associated with the use of an urinary catheter (100%, P < 0.001), previous surgery (44%, P < 0.05) and trachostomy (40%, P < 0.05), when compared with those of non-ICU patients. CONCLUSIONS: The identical RAPD pattern of all C. tropicalis strains from ICU patients indicates that they possibly originated from one clone. Through the investigation of risk factors, we can postulate that an urinary catheter might be a source for the outbreaks of urinary tract infections in the ICU. In addition, RAPD analysis might be a very useful test as one of the molecular epidemiologic tools for C. tropicalis funguria.