Theoretical analyses have shown that the spatial scaling of environmental autocorrelation, strength of density regulation, and the dispersal of individuals determine the scaling of synchrony in population fluctuations. By modeling the separate effects of density regulation, environmental stochasticity, and demographic stochasticity, we estimate the spatial scaling of the component that is due to environmental stochasticity in the population dynamics of roe deer (Capreolus capreolus) in Norway. The estimated spatial scaling of the environmental noise was similar to 200 km. An examination of how different weather variables influenced the scaling indicated that snow depth was the major weather variable affecting the scaling of synchrony in population fluctuations, and was negatively related to population growth rates in 97.4% of the 151 populations included in the study. A large-scale climatic phenomenon, the North Atlantic Oscillation, was positively related to population growth rates in 94.7% of the populations but did not significantly affect the pattern of synchrony among populations. We used newly developed theoretical results of the contribution of environmental noise and dispersal to the spatial scale of synchrony to show that the spatial scaling estimated in this study could not be explained by dispersal. This suggests that common environmental noise operating mainly during the winter is able to synchronize population fluctuations of roe deer over large distances.