Few studies have examined the sources of variability in cardiac function measurements in unrestrained animals and the impact of this variability on detection of treatment effects. The heart rate was monitored with implanted ECG transmitters in two groups of male rats, age 7 and 23 mo. Animals were monitored in their cages to determine optimal heart rate sampling frequency and sources of variability in heart rate, including whether there were persistent animal-to-animal differences. Ambient temperature was transiently increased to test whether correction for animal-to-animal differences improved sensitivity for detection of treatment effects. Animal-to-animal differences were statistically significant and accounted for about 18.3% and 11.5% of the total variance for old and young rats, respectively. In both the old and young rats, the heart rate decreased during the heat challenges relative to the control group, but the noncorrected differences were not statistically significant. When pre-exposure baseline values for each rat (average of 72 h prior to the first temperature challenge) were subtracted, the decrease in heart rate was statistically significant during all three challenges for both old and young rats. Subtraction of preexposure heart rate data to correct for baseline differences between animals is important for measuring treatment effects.