Corridors have become a key issue in the discussion of conservation planning; however, few empirical data exist on the use of corridors and their effects on population dynamics. The objective of this replicated, population level, capture-recapture experiment on meadow voles was to estimate and compare population characteristics of voles between (1) corridor-linked fragments, (2) isolated or nonlinked fragments, and (3) unfragmented areas. We conducted two field experiments involving 22600 captures of 5700 individuals. In the first, the maintained corridor study, corridors were maintained at the time of fragmentation, and in the second, the constructed corridor study, we constructed corridors between patches that had been fragmented for some period of time. We applied multistate capture-recapture models with the robust design to estimate adult movement and survival rates, population size, temporal variation in population size, recruitment, and juvenile survival rates. Movement rates increased to a greater extent on constructed corridor-linked grids than on the unfragmented or non-linked fragmented grids between the pre- and post-treatment periods. We found significant differences in local survival on the treated (corridor-linked) grids compared to survival on the fragmented and unfragmented grids between the pre- and post-treatment periods. We found no clear pattern of treatment effects on population size or recruitment in either study. However, in both studies, we found that unfragmented grids were more stable than the fragmented grids based on lower temporal variability in population size. To our knowledge, this is the first experimental study demonstrating that corridors constructed between existing fragmented populations can indeed cause increases in movement and associated changes in demography, supporting the use of constructed corridors for this purpose in conservation biology.