Northern high-latitude regions could feed back strongly on global warming because of large carbon pools and the fact that those regions are predicted to experience temperature increases greater than the global average. Furthermore, ecological functioning and carbon cycling are both strongly related to the prevailing hydrological conditions. In this study, we address these issues and present a newly developed model LPJ distributed hydrology (LPJ-DH) with distributed hydrology based on the dynamic global ecosystem and biogeochemistry model LPJ-GUESS. The new model is an enhanced version of LPJ-GUESS, introducing parametrizations of surface water routing and lateral water fluxes between grid cells. The newly introduced topographic variables in LPJ-DH are extracted from digital elevation models. LPJ-DH is tested at a 50-m resolution in the Stordalen catchment, northern Sweden. Modelled runoff is evaluated against the measured runoff from 2007 to 2009 at six outlet points. We demonstrate that the estimated monthly runoff from LPJ-DH agrees more closely with the measured data (adjusted R2=0·8713) than did the standard LPJ-GUESS model (adjusted R2=0·4277). However, there are still difficulties in predicting low-flow periods. The new model shows a possible advantage in representing the drainage network as well as topographic effects on water redistribution. The modelled birch tree line is in the range of the imagery observation, and the model captures the observed values of vegetation biomass in the region. Significant changes in biomass and carbon fluxes are also observed in the new model. Generally, the study justifies the feasibility and advantages of incorporating distributed topographic indices into the dynamic ecosystem model LPJ-GUESS. © 2013 John Wiley & Sons, Ltd.