Abstract. Increasing rates of biodiversity loss are adding momentum to efforts seeking to restore or rewild degraded landscapes. Here, we investigated the effects of natural forest regeneration on water flux partitioning, water ages and hydrological connectivity, using the tracer-aided ecohydrological model EcH2O-iso. The model was calibrated using ~3.5 years of diverse ecohydrological and isotope datasets available for a catchment in the Scottish Highlands, an area where the impetus for regeneration of native pinewoods is growing. We then simulated two land cover change scenarios that incorporated forests at early (thicket) and late (old-open forest) stages of regeneration, respectively, and compared these to a present-day baseline simulation. Changes to forest structure (proportional vegetation cover, vegetation heights and leaf area index of pine trees) were modelled for each stage. Establishment of thicket forest had the greatest effect on water partitioning/ages and connectivity, with increased losses to interception evaporation driving reductions in below-canopy fluxes (soil evaporation, groundwater recharge and streamflow) and generally slower rates of water turnover. Effects on streamflow were most evident for low and moderate summer flows rather than winter high flows. Whilst full forest regeneration was limited to hillslopes, resultant changes to the spatial dynamics of flux partitioning could also cause drying out of the valley bottom. The more open nature of the older forest generally resulted in water fluxes, ages and connectivity characteristics returning towards baseline conditions. Our work implies that the ecohydrological consequences of natural forest regeneration on degraded land depend on the structural characteristics of the forest at different stages of development. Consequently, future land cover change investigations need to move beyond consideration of simple forest vs. non-forest scenarios to inform management that effectively balances landscape restoration with demand for ecosystem services. Tracer-aided ecohydrological models were also shown to be useful tools for land cover change simulations and further potential of such models was highlighted.