PURPOSE: The purpose of study was to evaluate the reliability of transrectal ultrasonography (TRUS) in the prostate volume measurement. MATERIALS AND METHODS: Forty-one patients underwent two TRUS examinations during two independent sessions to measure the three-dimensional diameters and volume of the prostate. The difference between the two TRUS examinations was evaluated, and the reliability of TRUS was assessed using the intraclass correlation coefficient (ICC), coefficient of variation (CV), and Bland and Altman plot. RESULTS: The patients were divided into two subgroups by two observers those with the mean prostate volume same or greater than 30 ml (group A) and those with the mean prostate volume less than 30 ml (group B). A prostate volume difference greater than 5 ml and 10 ml between the two TRUS examinations was more frequently included in the group A than in the group B (43% and 33%, 5% and 0%, respectively) (p < 0.05). ICC and CV showed poor reliability for measuring the prostate height, length and volume in the group A (ICC for width, height, length and volume < 0.75, CV for volume =12.97%). CONCLUSION: Prostate volume measurement by TRUS may be unreliable in large prostates with unreliable measurement of the height and length being the primary limiting cause.