The effects of habitat destruction and fragmentation on survival rate and population size were tested in an experimental model system consisting of seven enclosed populations of root voles, Microtus oeconomus. The study comprised two consecutive parts. First, the effect of habitat fragmentation was investigated by comparing survival in three populations inhabiting continuous habitat with that in for populations inhabiting fragmented habitat. Second, the continuous habitats were experimentally fragmented by destroying 50% of the area inhabited by the populations, producing the same configuration as in the fragmented habitats. These extensive habitat manipulations were expected to affect survival rates due to changes in the spatiosocial structuring of the populations. Throughout the study, populations were monitored by an intensive live—trapping program. Cormack—Jolly—Sebver methodology was used for statistical modelling and estimation of survival rates. The seven experimental populations differed considerably in population dynamics: some grew, while others remained remarkably stable throughout the experiment. In contrast to expectations, neith population size nor survival rates seemed to be affected by the experimental treatments. survival rate seemed to play a minor role in determining population size. Emigration rates, which also seemed to have little direct demographic importance, were considerably higher in fragmented than in continuous habitats during the first (pre—destruction) part of the study. Surprisingly, this difference persisted after habitat destruction, when all populations experienced the same habitat configuration. We conclude that survival rates in these root vole populations were not dependent on major alterations of spatiosocial structure, possibly owing to adaptation to natural disturbance regimes.