Time series data on cyclic vole populations in 1971–1999 in northern Sweden show a long-term decline in numbers from the early 1970s to the late 1990s. We tested the destructive sampling hypothesis predicting that previously unsampled (new) plots in remote areas would yield higher density indices than the density level of permanent plots in the early 1970s. In autumn 1999 we sampled both permanent (treatment) plots and new (control) plots. Density was not higher on new than on permanent sampling plots for any of the predominant vole species, Clethrionomys glareolus, C. rufocanus, and Microtus agrestis. It appears unlikely that destructive sampling has caused the observed decline in vole numbers.