Purpose: The purpose of this study was to examine the relationship among helmet surface temperature (Thl), head top temperature (Thd) and tympanic temperature (Tty) during American football practice in summer. Methods : The subjects were collegiate American football players. Temperatures were measured in August in 1993 and 1994. Thl, Thd and Tty were measured by infrared tympanic thermometers. Environmental temperatures that were measured were dry-bulb temperature (Td), wet-bulb temperature (Tw), globe temperature (GT) and wet-bulb globe temperature (WBGT) . Results: Significant correlations were observed among all measured temperatures (P<0.0001) . High coefficients of correlation were observed among Thl, Thd and GT. The highest relationships were observed between Thl and Thd (r=.727), and between Tty and Td (r=.766) . The coefficient of correlation between Tty and Thl was higher than that between Tty and Thd. Heat stress of the whole body (F1; heat stress factor: Tty, Td, Tw and WBGT, proportion=71.4%) and head environment factor (F2 ; helmet factor : Thl, Thd and GT, proportion=14.3%) were chosen in factor analysis. A close relationship was observed between both factors (r=.773) . Both GT and Thl showed a high correlation with Thd, which suggests the influence of radiant heat through a helmet on the whole body. Conclusion : The temperature in a helmet is a micro environment temperature surrounding the head. Accordingly, the heat load is reduced by taking the helmet off frequently during football practice.