A procedure for estimating urine volumes from urine weights is presented. Regression analysis was used to assess correlation between weight-volume errors. The urine weight was divided by 1.016, the mean urine specific gravity in our patient population. It was found that the gravimetric method gave results lower than the volume method by one percent or less. We conclude that this difference is acceptable for clinical purposes. The gravimetric procedure has the advantages of rapidity, cost-effectiveness, and safety.