Live and kill traps were used to remove House mice living in granaries, mixed-food stores, dairy units and ancillary stores on farms. Structural weaknesses were common to all buildings. Cavities in walls, in particular, provided favourable living and nesting quarters for mice and were rapidly exploited. The body-weight structure and reproductive traits of the populations in the different habitats were compared. The onset of maturity, in both sexes, was closely related to body-weight increase in the 10 to 14 g range. Breeding, unlike that in most out-door populations in temperate climates, continued throughout the year. Ancillary buildings holding limited food supported small reservoir populations. Mice inhabiting dairy units containing pelleted cattle feed tended to be heavier than those sampled elsewhere and females in the units produced the largest litters. Reproductive performance on farms was intermediate between that of mice occupying urban buildings and corn ricks. There was evidence that environmental and behavioural factors were operating to limit increase in numbers in the dairy units.