Live-trapping studies at the Patuxent Wildlife Research Center, Maryland, showed that the ranges of wood mice were larger when the population density was lower and smaller when the population density was higher. When the population density was about 1.3 male mice per acre in June 1954, the average distance recorded between traps after four or more captures was 258 feet. When the population density was about 4.1 male mice per acre in June 1957, the average distance was 119 feet. Differences were statistically significant. Females were so scarce at the low that comparisons could not be made for them. Examples from the literature also show that home range of a species may vary with population density. Other examples show that the range may vary with habitat, breeding condition and food supply. These variations in range size reduce the reliability of censuses in which relative methods are used: Lines of traps sample the population of a larger area when ranges are large than they do when ranges are small. Direct comparisons therefore will err in some degree. Error may be introduced also when line-trap data are transformed to per acre figures on the basis of home-range estimates made by area-trapping at another place or time. Variation in range size also can make it necessary to change area-trapping plans, for larger quadrants are needed when ranges are larger. It my be necessary to set traps closer together when ranges are small than when ranges are large.