Populations of Wood mice Apodemus sylvaticus and Bank voles Clelhrionomys glareolus were studied by mark-recapture techniques in five different woodlands but concentrated on an oak Quercus petraea wood and a yew Taxus baccata wood, where food supply was also investigated. Much of the yew wood was on limestone pavement which was greatly creviced thus providing static and quantifiable cover. Indices of the intensity with which food was being searched for were obtained on all areas by seed-removal experiments.

Trappability varied within the populations and seasonally, being highest in summer. Unusually high densities of mice were recorded in the yew wood but not elsewhere; vole densities were always low. There was an inverse relationship between range length and density: mean range length in the yew wood was exceptionally low. Males generally ranged further than females, voles further than mice. Females outnumbered males in the oak wood, where cover was minimal, when the population was low. In the two areas with most cover, males outnumbered females. In a further area, where numbers were initially exceedingly low, the population consisted entirely of males for almost a year. Sexes differed in habitat preference but numbers, especially of males, were significantly correlated with cover. Voles usually bred for shorter periods than mice. Overwinter breeding and increased overwinter survival followed good seed years. Individuals recruited when food was plentiful lived longest. In seed-removal experiments rate of removal increased with shortage of natural seed.

Food and intraspecific aggression were the probable major factors in regulating population size, of mice at least. Cover probably modifies intrinsic mechanisms by reducing numbers of encounters between individuals.