Intermittent rivers and ephemeral streams (IRES) account for more than half of the world’s river networks and are considered to increase under climate change and growing anthropogenic water use. However, the hydrological mechanisms that control the spatio-temporal flow patterns in IRES and their effects on the expansion and contraction of stream segments are not fully understood. Aiming to overcome this problem, DRYvER developed a hybrid hydrological modeling approach to simulate daily flow conditions and provide flow intermittence indicators in 6 European Drying River Networks (DRNs). The JAMS/J2000 process-based hydrological model forced with the ERA5-land reanalysis is used to simulate daily discharges at high spatial resolution, which includes the scale of small reaches (about 50 ha catchment size). As the JAMS/J2000 does not allow to reproduce accurately the periods with no flow, a Random Forest classification model (RF) was used to predict the flow condition (dry or flowing) at the reach scale in the DRNs, using hydrological variables simulated with JAMS/J2000 as explanatory variables, and observed flow intermittence data (e.g. data from field measurements, citizens science application) to train the RF model. Results show that the hybrid hydrological modeling approach (JAMS/J2000 and RF) enables to reproduce accurately the spatio-temporal patterns of flow intermittence in the Albarine (France), Bukkosdi (Hungary), Lepsämänjoki (Finland), and Velička (Czech Republic) DRNs. Additional observed flow intermittence data would be needed in the Genal (Spain) DRN in order to improve the simulation results. The prediction of flow conditions could not be produced for the Butižnica (Croatia) DRN due to a lack of observed flow intermittence data.