The HBV model, a conceptual rainfall-runoff model, was applied to 11 catchments within the NOPEX area. The catchment areas ranged from 7 to 950km² with between 41 and 87% covered by forest. The aim was to relate the different model parameters to physical catchment characteristics. Such relationships would allow simulating runoff from ungauged catchments and could be used to discuss the physical basis of the model. Using a 9-year calibration period the best parameter sets were determined for each catchment. A Monte Carlo procedure and two different criteria were used for the optimisation: the common efficiency and a fuzzy measure that combined different objective functions and was found to reduce parameter uncertainty. The runoff simulations of the model agreed well with the observed runoff series and relationships to catchment characteristics could be found for six of the 13 parameters. The goodness of runoff predictions using derived regional parameter sets was tested with variable results. Some relationships between lake percentage and soil parameters called the physical basis of the model into question as they could not be explained by the physical processes in the soil but by the dominating effect of lakes to runoff variations. On the other hand, relationships between forest percentage and snow parameters supported the physical basis of the model.