PURPOSE: To assess the usefulness of implanted fiducial markers in the setup of hypofractionated radiotherapy for prostate cancer patients by comparing a fiducial marker matched setup with a pelvic bone match. MATERIALS AND METHODS: Four prostate cancer patients treated with definitive hypofractionated radiotherapy between September 2009 and August 2010 were enrolled in this study. Three gold fiducial markers were implanted into the prostate and through the rectum under ultrasound guidance around a week before radiotherapy. Glycerin enemas were given prior to each radiotherapy planning CT and every radiotherapy session. Hypofractionated radiotherapy was planned for a total dose of 59.5 Gy in daily 3.5 Gy with using the Novalis system. Orthogonal kV X-rays were taken before radiotherapy. Treatment positions were adjusted according to the results from the fusion of the fiducial markers on digitally reconstructed radiographs of a radiotherapy plan with those on orthogonal kV X-rays. When the difference in the coordinates from the fiducial marker fusion was less than 1 mm, the patient position was approved for radiotherapy. A virtual bone matching was carried out at the fiducial marker matched position, and then a setup difference between the fiducial marker matching and bone matching was evaluated. RESULTS: Three patients received a planned 17-fractionated radiotherapy and the rest underwent 16 fractionations. The setup error of the fiducial marker matching was 0.94+/-0.62 mm (range, 0.09 to 3.01 mm; median, 0.81 mm), and the means of the lateral, craniocaudal, and anteroposterior errors were 0.39+/-0.34 mm, 0.46+/-0.34 mm, and 0.57+/-0.59 mm, respectively. The setup error of the pelvic bony matching was 3.15+/-2.03 mm (range, 0.25 to 8.23 mm; median, 2.95 mm), and the error of craniocaudal direction (2.29+/-1.95 mm) was significantly larger than those of anteroposterior (1.73+/-1.31 mm) and lateral directions (0.45+/-0.37 mm), respectively (p<0.05). Incidences of over 3 mm and 5 mm in setup difference among the fractionations were 1.5% and 0% in the fiducial marker matching, respectively, and 49.3% and 17.9% in the pelvic bone matching, respectively. CONCLUSION: The more precise setup of hypofractionated radiotherapy for prostate cancer patients is feasible with the implanted fiducial marker matching compared with the pelvic bony matching. Therefore, a less marginal expansion of planning target volume produces less radiation exposure to adjacent normal tissues, which could ultimately make hypofractionated radiotherapy safer.