We studied rectal body temperatures of house mice (Mus domesticus) that had been artificially selected for high voluntary wheel running.1. At generation 17, mice from the four replicate selected lines ran, on average, 2.5-times as many revolutions/day as did mice from the four random-bred control lines.2. During the day, repeatability of individual differences in body temperature measured 4 days apart was low; at night, repeatability was statistically significant across three time scales (1 day, 1 week, 2 weeks).3. During the day, body temperatures of selected and control animals did not differ; at night, mice from selected lines had higher body temperatures. However, when amount of wheel running immediately prior to measurement was included as a covariate, the difference was no longer statistically significant.Higher body temperatures, associated with increased activity, might enhance locomotor abilities through Q10 effects, increase metabolic rate and food requirements, affect sleep patterns, and alter expression of heat-shock proteins.