OBJECTIVE: The aim of this study was to evaluate the utility of combined positron emission tomography/computed tomography (PET/CT) for identifying ovarian cancer tumor masses > or =1 cm in patients with clinically occult recurrent disease by conventional CT imaging. METHODS: Twenty-two patients with epithelial ovarian cancer, rising serum CA125 levels, and negative or equivocal conventional CT imaging > or =6 months after primary therapy underwent combined PET/CT imaging followed by surgical reassessment. Fisher's exact test was used to measure the ability of PET/CT to predict macroscopic disease > or =1 cm. RESULTS: The median patient age was 55 years, and 91% of patients had FIGO Stage IIIC/IV disease. The median increase in serum CA125 was 24 U/ml (range 10 to 330 U/ml). Conventional CT was reported as negative (n = 15) or equivocal (n = 7) in all cases. Eighteen patients were ultimately found to harbor recurrent ovarian cancer measuring > or =1 cm at the time of surgery, with a median maximal tumor diameter of 2.3 cm (range 1.5 to 3.2 cm). The overall patient-based accuracy of PET/CT in detecting recurrent disease > or =1 cm was 81.8%, with a sensitivity of 83.3% and positive predictive value of 93.8% (P = 0.046). Of patients with recurrent ovarian cancer > or =1 cm, complete cytoreduction to no gross residual tumor was accomplished in 72.2%. CONCLUSION: PET/CT imaging demonstrates high sensitivity and positive predictive value in identifying potentially resectable, macroscopic recurrent ovarian cancer among patients with biochemical evidence of recurrence and negative or equivocal conventional CT findings. In appropriately selected patients, early identification of macroscopic recurrent disease may facilitate complete surgical cytoreduction.