This study investigated the feasibility of remotely estimating the urinary flow velocity of a human subject with high accuracy using millimeter-wave radar. Uroflowmetry is a measurement that involves the speed and volume of voided urine to diagnose benign prostatic hyperplasia or bladder abnormalities. Traditionally, the urine velocity during urination has been determined indirectly by analyzing the urine weight during urination. The maximum velocity and urination pattern were then used as a reference to determine the health condition of the prostate and bladder. The traditional uroflowmetry comprises an indirect measurement related to the flow path to the reservoir that causes time delay and water waves that impact the weight. We proposed radar-based uroflowmetry to directly measure the velocity of urine flow, which is more accurate. We exploited Frequency-Modulated Continuous-Wave (FMCW) radar that provides a range-Doppler diagram, allowing extraction of the velocity of a target at a certain range. To verify the proposed method, first, we measured water speed from a water hose using radar and compared it to a calculated value. Next, to emulate the urination scenario, we used a squeezable dummy bladder to create a streamlined water flow in front of the millimeter-wave FMCW radar. We validated the result by concurrently employing the traditional uroflowmetry that is based on a weight sensor to compare the results with the proposed radar-based method. The comparison of the two results confirmed that radar velocity estimation can yield results, confirmed by the traditional method, while demonstrating more detailed features of urination.