Background: Positron emission tomography (PET) with F-18 fluorodeoxyglucose (FDG) is a useful technique for detection of malignancy. The purpose of this study was to determine if FDG-PET scanning using visual and quantitative analyses can identify patients with recurrent colorectal tumor following abdominoperineal resection. Methods: Eighteen patients were evaluated for possible local recurrence of rectal carcinoma following abdominoperineal resection. The clinical presentation included rising carcinoembryonic antigen levels (n = 5), increasing size of a presacral mass on computed tomography or magnetic resonance (n = 13), or local symptoms (n = 3). Axial PET images of the pelvis were obtained following an injection of 10 mCi of FDG prior to biopsy. Quantitative analysis was performed by calculation of a standardized uptake ratio (SUR), and the images were interperated by two radiologists in consensus. FDG-PET findings were correlated with histological or cytological findings and with the clinical outcome. Results: Recurrent malignancy was confirmed in 13 patients by surgery (n = 8) or percutaneous biopsy (n = 5). Benign lesions were confirmed in five patients by surgery (n = 1), biopsy (n = 3), or clinical follow up (n = 1). Visual analysis of the FDG-PET data had a sensitivity of 92.3% (12/13) for recurrent disease (95% confidence limits; 63.9%, 99.8%) and a specificity of 80% (4/5; 95% confidence limits; 28.3%, 99.4%). SUR values were significantly higher in malignant lesions (range = 2.92–19.74, mean = 6.89) than in benign ones (range = 1.40–3.47, mean = 1.96; p= 0.002). Conclusion: FDG-PET is an accurate technique for detection of locally recurrent colorectal carcinoma. Visual analysis is equivalent to quantitative analysis for detection of disease.