OBJECTIVES: The most precise procedure for the measurement of residual urine volume is bladder catheterization; nevertheless, it takes real associated morbidity. Our objective is to define the accuracy of bladder ultrasonography when residual volume estimation is performed. METHODS: Thirty patients were studied. Ultrasound estimation of residual volume was undertaken with Simpson's method V = (pi/6) x L x T x AP. Obtained values were contrasted with the post-catheterization registered ones. RESULTS: The mean ultrasound estimated residual urine volume was 82.1 ml., versus 156.5 after catheterization. The mean difference was 74.4 ml. (significant, p < 0.001). Correlation coefficient was 0.93. Regression equation was calculated, a well as the corrected formula that, according to the outputs of our study, offers greater reliability in the estimate of the aforementioned volume. Ultrasound sensitivity when proper estimation of volumes equal or greater to 100 cc. is carried out was only 47%, opposite to the elevated specificity (100%) for this purpose. CONCLUSIONS: Transabdominal ultrasound underestimates, to a large extent, the values of residual urine volume. We consider that ultrasound estimation of residual volume constitutes an alternative to be kept in mind; nevertheless, the formula to calculate this volume through bladder diameters introduces an important error and so, it should be corrected.