We show that the large spatial aggregation of model parameters in common catchment scale nitrogen budget modeling leads to artifacts that may, for instance, be a factor in explaining reported decreases of calibrated in-stream nitrogen loss rates, λs*, with increasing stream size. In general, the common assumption of a single representative solute travel time for an entire stream reach may lead to considerable underestimation of the actual underlying local biogeochemical loss rate λs by λs*, which increases with actual λs value and/or increasing mean solute travel time and travel time variability in the stream. We propose an up-scaling methodology to overcome such model artifacts, in form of closed-form expressions of catchment-scale, in-stream nitrogen delivery factors for diffuse and point sources, as functions of local-scale nitrogen loss rates, λs.