In order to verify the impact of riverbed measurement using the Acoustic Doppler Current Profiler (ADCP) based on the Doppler effect on the accuracy of the structure of water surface level simulated in the numerical model, two series of field measurements were performed in the Skawa River, and another series in a hydraulic laboratory. The reference measurements of gravel riverbed were conducted using the classical method, by applying a measuring device. Precise measurement of riverbed elevations in the field was possible thanks to the use of a portable stable measuring station, which ensures stable probe movement during measuring. The obtained results were evaluated using the t-Student test, as well as the methods proposed by Ozga-Zielińska with Brzeziński, Moriasi, and Legates with McCabe. The analysis of the conformity between the results obtained using the ADCP methods and the reference measurements showed very good compatibility in the representation of riverbed elevations for the RSR, NSE and PBIAS evaluation statistics in the case of the laboratory series. The values obtained in both field series revealed an unsatisfactory representation of the riverbed elevations of the RSR and NSE evaluation statistics. Differences in the riverbed level have translated into differences in the water surface level. For the discharge of Q_1%, the said differences do not exceed 0.066 m, which corresponds to the mean diameter d_m of bed material. Measurement of the riverbed configuration using the ADCP affects the water surface level and thus also the flood hazard zone. In mountainous areas, where the depth differences of the riverbeds and floodplains are significant, this impact is limited.