Seasonality has been suggested as a necessary factor for the initiation of vole population cycles in Fennoscandia. This has been well described for a latitudinal gradient. Here, we used an elevational gradient as a proxy for winter length to study how the length of the winter season correlates with the amplitude of bank vole Myodes glareolus population cycles. In addition, we studied whether the small mammalian generalist predator community present locally could explain any elevational effects. We estimated the population size of 30 local bank vole populations. We found a strong effect of elevation on the amplitude of the population cycle with local populations at around 800 m elevation having 1.5 times greater densities than populations in the valley (ca 300 m elevation). A univariate model with elevation as predictor for amplitude was twice as likely to be the best model than models including generalist predators. Our results fit well with the theories of a positive effect of winter length on the amplitude of vole population cycles in Fennoscandia, irrespective of whether the seasonal effect corresponds to a latitudinal or elevational gradient. The mechanisms may be limited resources during winter rather than generalist predators.