A combined simulation of runoff dynamics and distributed transport of the stable isotope oxygen-18 (18O) was performed in the mountainous Dreisam basin (258 km2) in the southwest of Germany. For this purpose, a mixing cell solute routing scheme was implemented in the fully distributed, process-orientated catchment model TACd. Perfect mixing in the catchment reservoirs was assumed in order to keep the model simple. In addition, effects like fractionation, diffusion into and out of immobile phases, as well as kinematic effects were neglected. Although results show a generally good agreement between simulated and measured concentrations of 18O in stream discharges, we discovered that the initial objective of validating the process representation of TACd by incorporating conservative solute transport cannot be achieved with the given data. Simulation misfits cannot be clearly attributed to process descriptions, mainly due to large uncertainties in the input regionalization of precipitation and 18O, and due to the lack of data for model initialization. Nevertheless, we critically evaluate the model’s conceptualization and parameterization, in order to deliver insight to the potential and limitations of distributed modelling of 18O signals on the meso-scale. In addition, the large influence of surface runoff generated on a small fraction of the total catchment area, and fast interflow components on the 18O values in total stream discharge could be demonstrated for the investigated events. This paper documents the first attempt to simulate the distributed 18O balance in a mesoscale catchment with temporal resolution on an hourly basis in order to optimize future measurement campaigns and modelling attempts.