Mass changes in the hydrosphere represent an important contributor to polar motion (PM) variations, especially at seasonal time scales (i.e., annual and semiannual). Although well studied, hydrological angular momentum (HAM) remains a major source of uncertainty in estimating PM excitation. In this work, we use a large number of climate models from the sixth phase of the Coupled Model Intercomparison Project (CMIP6) to determine HAM series both from individual models and their combination, formed with a multi-model mean, a weighted mean, and a three-cornered hat (TCH) method. The CMIP6-based HAM series are analysed in several spectral bands and evaluated with a reference hydrological signal in geodetically observed PM excitation (GAO). HAM determined from CMIP6 were also compared to HAM calculated from Gravity Recovery and Climate Experiment (GRACE) measurements. We find that while climate models do not allow for reliable estimation of non-seasonal changes in HAM, they can help interpret seasonal variability. For annual prograde and semiannual retrograde oscillations, several combined CMIP6-based series exhibit higher amplitude and phase consistency with GAO than the corresponding series computed from GRACE data. Whether one uses a simple average of the models, a weighted average, or a combination of models from the TCH method has little impact on the resulting HAM series and their level of agreement with GAO. Our study advances the understanding of hydrological signal in Earth’s rotation at seasonal time scales.