PURPOSE: To evaluate the accuracy of [(18)F]-choline (FCH) positron emission tomography/computed tomography (PET/CT) for staging and restaging of prostate cancer. METHODS: FCH PET/CT was performed in 111 patients with prostate cancer using 200 MBq FCH: 43 patients [mean age 63 years; mean prostrate specific antigen (PSA) 11.58 microg/l] were examined for initial staging, and 68 patients (mean age 66.4 years) were examined for restaging (mean PSA 10.81 microg/l). FCH PET/CT results were correlated to histopathology, bone scan, morphology as revealed by magnetic resonance imaging (MRI) and CT, PET/CT follow-up and PSA follow-up after therapy. RESULTS: FCH PET/CT scans at initial staging correctly showed no metastases in 36/38 patients undergoing radical surgery, as confirmed by PSA levels <0.1 microg/l 6 months postoperatively. Lymphadenectomy was performed in 24 of these patients, revealing four false FCH-negative lymph nodes (LN). In one patient, only lymphadenectomy was performed since a FCH-positive LN was confirmed by histology. Four patients showed FCH-positive bone metastases, as proven by bone scan. FCH PET/CT scans at restaging correctly revealed local recurrence in 36 patients. No pathological FCH uptake was observed in 11 patients with biochemical recurrence. Twenty-three patients showed FCH-positive LN. Twenty LN were surgically removed in seven patients. Histopathology verified metastases in all LN, but revealed two additional metastastic, FCH-negative LN. Seventeen patients showed FCH-positive bone metastases, as proven by bone scan or MRI. Sensitivity to detect recurrent disease was 86%. CONCLUSION: The results obtained using FCH PET/CT scans for initial N-staging were discouraging, especially in terms of its inability to detect small metastases. Recurrent disease can be localized reliably in patients with PSA levels of >2 microg/l.