PURPOSE This report concentrates on the localization of the prostatic apex using implantable markers, and a comparison to localization defined by computed tomography (CT) and retrograde urethrography. METHODS AND MATERIALS Fifteen patients were entered into a prospective trial and scheduled to undergo (a) pelvic CT, (b) retrograde urethrogram, and (c) transrectal ultrasound with placement of radiodense markers. Three markers were implanted: one was placed at the trapezoid area (prostatic apex), and the others placed lateral to the base of each seminal vesicle. The retrograde urethrogram was performed using standard technique. The superior-inferior distance between the apex identified by the marker placed at the prostatic apex and the other studies was measured. RESULTS CT and urethrogram overestimated the inferior extent of the prostatic apex when compared to the location as defined by the implanted marker. With CT, the average distance from the marker to the CT-defined apex was 0.6 cm (95% C.I.--0.4-0.8 cm). With urethrogram, the average distance from the marker to the urethrogram-defined apex was 1.3 cm (95% C.I.--0.7-1.9 cm). When CT and urethrogram were compared, CT was more accurate in identifying the prostatic apex. CONCLUSION Under ultrasound guidance, radiodense markers have been implanted into the prostate. This has revealed that the apex is localized superior to the apical margin as defined by retrograde urethrogram and CT, and that CT may be more accurate than retrograde urethrogram. In addition, the placement of multiple markers yields spatial information on prostatic position that can be extracted from megavoltage portal images.