To compare the visibility of anatomical details and prostate cancer local staging performance of pelvic phased-array coil and integrated endorectal-pelvic phased-array coil MR imaging, with histologic analysis serving as the reference standard. MR imaging was performed in 81 consecutive patients with biopsy-proved prostate cancer, prior to radical prostatectomy, on a 1.5T scanner. T2-weighted fast spin echo images of the prostate were obtained using phased-array coil and endorectal-pelvic phased-array coils. Prospectively, one radiologist, retrospectively, two radiologists and two less experienced radiologists working in consensus, evaluated and scored all endorectal-pelvic phased-array imaging, with regard to visibility of anatomical details and local staging. Receiver operator characteristics (ROC) analysis was performed. Anatomical details of the overall prostate were significantly better evaluated using the endorectal-pelvic phased-array coil setup (P<0.05). The overall local staging accuracy, sensitivity and specificity for the pelvic phased-array coil was 59% (48/81), 56% (20/36) and 62% (28/45), and for the endorectal-pelvic phased-array coils 83% (67/81), 64% (23/36) and 98% (44/45) respectively, for the prospective reader. Accuracy and specificity were significantly better with endorectal-pelvic phased-array coils (P<0.05). The overall staging accuracy, sensitivity and specificity for the retrospective readers were 78-79% (P<0.05), 56-58% and 96%, for the endorectal-pelvic phased-array coils. Area under the ROC curve (Az) was significantly higher for endorectal-pelvic phased-array coils (Az=0.74) compared to pelvic phased-array coil (Az=0.57), for the prospective reader. The use of endorectal-pelvic phased array coils resulted in significant improvement of anatomic details, extracapsular extension accuracy and specificity. Overstaging is reduced significantly with equal sensitivity when an endorectal-pelvic phased-array coil is used.