In adult humans, core temperature is influenced by activity; the sensitivity of core temperature to such effects shows a phase dependence and is also influenced by the environment and whether the individual is asleep or awake. We have investigated if similar effects are evident in neonates, in whom thermoregulation and the circadian rhythm of core temperature are not fully developed. Eleven full-term, healthy babies were studied singly (light 07:00-19:00) at 2 days of age and again 4 weeks after birth; between these times, they were tended routinely on a communal ward. On study days, 10-minute recordings were made of rectal and skin (abdominal) temperature, heart rate (HR), and behavioral state. Sensitivities of the temperatures to activity ('arousal') were assessed throughout the 24h by measuring the gradient of (temperature/HR). Sensitivities measured at 01:00, 05:00, 09:00, 13:00, 17:00, and 21:00 were used as dependent variables in stepwise regression and linear regression analyses, with 'subjects,' 'light versus dark,' 'behavioral state,' and 'difference between time of measurement and the acrophase of the endogenous component of the temperature rhythm' (ignoring sign) as possible predictors. (Acrophases of the temperature rhythms had been estimated from 24h data purified using the behavioral state record.) Light versus dark acted as a significant predictor of the sensitivity of rectal temperature to arousal on day 2 and week 4, the sensitivity increasing in the light, and there was limited evidence for behavioral state acting as a predictor on day 2. Neither factor was a significant predictor when the sensitivity of the babies' skin temperatures to arousal was investigated. There was also some evidence that the difference between the time of measurement and the temperature acrophase acted as a predictor of sensitivity to arousal in both rectal (day 2) and skin (week 4) temperature, with larger differences decreasing the sensitivity. These results indicate that there are masking effects on body temperature due to arousal in neonates, the size of which depends on both internal and external factors. However, this sensitivity of temperature to arousal shows differences from the sensitivity of temperature to physical activity in both adult humans and adult mice. One possible explanation of this result is that temperature regulation and the circadian system are not fully developed in humans at this age.