A population of meadow voles Microtus pennsylvanicus in Minnesota exhibiting annual rather than multiannual fluctuations in density was studied. To characterize voles of different movement patterns, data on sex, age, reproductive condition and genotype at two polymorphic loci (Tf arid LAP) were collected in summer (n = 1323) and autumn (n = 1191) during 2 consecutive years. Three movement patterns were identified: (1) dispersers were those voles that moved onto grids from which all trapped Microtus were removed; (2) residents were those voles found on the same grid during at least two trapping periods; (3) movers were those voles known to move long distances within a population. The following conclusions were drawn. (1) Dispersers differed from residents most consistently by being more likely to be young. Although dispersers also differed from residents both reproductively and genetically, the differences were not consistent for all age-sex groups throughout the study. (2) Movers were an identifiable subset of residents that differed from dispersers and other residents during summer by being more likely to be adult males. (3) During the breeding season, female residents were the most stable subset of the population, whereas adult males tended to move long distances within the population. Young females tended to disperse close to the time of initiating reproductive activity; young males tended to disperse as nonreproductives. During the nonbreeding season, dispersers were more nearly a random subset of residents. INTRODUCTION Dispersal, movement away from home resulting in at least temporary transience for an individual (Lidicker, 1975), can have both demographic and evolutionary effects in populations (for review see Gaines and McClenaghan, 1980). Evolutionarily, it affects gene flow and thus patterns of genetic change within and between populations (Anderson, 1970; Endler, 1977; Levin, 1976; Weins, 1976). In addition, the adaptive advantage of dispersal is unclear. Dispersers presumably encounter some or all of a multitude of risks inherent in traversing unfamiliar habitat. If dispersal is adaptive to individuals, then reproductive benefits such as outbreeding, finding new resources and avoiding aggression from conspecifics must on average outweigh the risks (Lidicker, 1962, 1975). Alternatively, dispersal behavior may not be adaptive to individuals; theoretically, it can evolve by parental manipulation (Hamilton and May, 1977) or by group selection (Van Valen, 1971). From the point of view of population biology, dispersal is of interest for two basic reasons. First, dispersal patterns may correlate with other life history characteristics (e.g., lifespan, age to first reproduction, and habitat, Pianka, 1970); thus, information on dispersal can aid in understanding the life history pattern of a population (Fairbairn, 1978). Secondly, dispersal can affect population density (Krebs and Myers, 1974; Lidicker, 1978; Finerty, 1980) and may alter the social structure of populations in such a way as to affect density (Charnov and Finerty, 1980). To test two hypotheses regarding dispersal in a population of Microtus pennsylvanicus, we established removal grids in both optimal and suboptimal habitat of a noncyclic population of M. pennsylvanicus in Minnesota. The objective of the research was to characterize voles of differing movement patterns (dispersers, movers and residents) with regard to sex, age, reproductive condition and genotype at the transferrin and 'Present address: BSCS, Box 930, Boulder, Colo. 80027. 2To whom reprint requests should be addressed.