A capture-mark-release study of deer mice (Peromyscus) was conducted on two 10-acre grassland areas (Zonal Air Pollution Systems or ZAPS) at monthly intervals from April to September 1976. Both areas were subdivided into four in-line experimental plots, with three smaller plots as buffer zones. Each of the experimental plots was fumigated at a median SO2 concentration of nearly 0 (control), 2.2 (low), 3.9 (medium) and 6.5 (high) pphm, respectively. Three functional groups of mice were recognized: resident adults and juveniles (group 1); transient adults and juveniles (group 2); and resident and transient adults (group 3). Temporal displacement and numerical replacement of group 1 was directed toward areas of lower SO2 concentration. Members of group 2, especially juveniles, were trapped more frequently on the control of low SO2 plots on ZAPS 2, but transient juveniles were virtually absent from ZAPS 1. Recruitment of group 3 onto the control plots at both ZAPS occurred more rapidly than onto other plots in late summer. Throughout most of the trapping period, the number of occupied traps on all fumigated plots decreased relative to control plots on both ZAPS, and remained relatively higher on control plots from mid-season to the experiments conclusion.