We developed and tested a regression model, interpolating long-term sequences of observed atmospheric deposition of SO42−, NO3−, and NH4+ in the upper Vltava river catchment (Czech Republic) to its three sub-regions, differing in elevation and forest cover. The model provides more realistic estimates of wet and total S and N depositions and their inter-annual variability in the study catchment than the available European deposition sequences, especially in the case of wet S deposition prior to 1997. In the model, ion fluxes are calculated as the product of the precipitation volume and ion concentrations, which both are derived as empirical functions of elevation and time. The long-term sequences of ion concentrations are based on measured deposition data at 19 stations and their relationships with central European emission trends of SO2, NOx, and NH3 for years with no measurements. Exponential relationships between elevation and precipitation volume (positive) and elevation and ion concentrations (negative) are used to convert the long-term sequences of precipitation and concentrations into values for individual elevations. Throughfall fluxes (TF) of S and N in the forest areas are calculated from their fluxes in precipitation (PF), using long-term sequences of TF:PF ratios, based on measured fluxes and the S and N emission trends. The calculated fluxes of S and reactive nitrogen (NO3–N and NH4–N) explain 80% and 56% of the variability in their measured fluxes, respectively, along an elevation gradient from 275 to 1334 m.