Global hydrological alterations are being driven by climate change. However, while modelling tools have been instrumental in identifying these changes, their calibration is often dependent solely on streamflow data, which limits the ability to simulate important hydrological processes with the required level of certainty. Constraining hydrological process simulations is required to identify robust model parameters, reducing model uncertainty and providing a baseline model for climate‐related risk assessments. In this study, stable isotope measurements of rainwater, groundwater and stream water (δ2H and δ18O), together with an end member mixing analysis (EMMA) (as a post‐processing step) were integrated with the Jena adaptative modelling system (JAMS)/J2000 rainfall‐runoff model, named J2000iso. The J2000iso models of δ2H and δ18O were able to achieve a streamflow Nash–Sutcliffe efficiency (NSE) of 0.75 and 0.72 and an isotope mixing Kling–Gupta efficiency (KGE) of 0.51 and 0.60, respectively. Compared with the base version, the J2000iso had a much smaller variability with 56% less ensemble variance in the NSE, 13% more simulated interflow which resulted in a better soil‐moisture simulation at an observation point, and lower simulated flow component uncertainty. The J2000iso δ2H model was more robust than the base version during a subsequent validation in terms of KGE and Bias with 0.60 and 0.002 compared with 0.51 and −0.14 for the base version. The EMMA provided a means to improve the hydrograph separation ability of the JAMS/J2000. As many catchments around the world are still ungauged and are affected by anthropogenic activities, the development and enhancement of isotope‐enabled modelling provides a means to improve the simulation of key hydrological processes.