Thermal responses and muscle performance in humans were studied during rest and exercise in a cool environment with different clothing distributions over the legs. Nine female subjects were exposed to 5 degrees C wearing shorts (SS), trousers with long legs (LL) or trousers with one long leg and one short leg (LS: LSc covered leg, LSu uncovered leg). The subjects also wore T-shirts and long-sleeved shirts. The subjects were seated for 60 min and after this they performed light stepping exercise for a further 60 min. Rectal temperature (T(re)) and skin temperature from seven (LL, SS) or nine sites (LS) were measured continuously. Surface electromyography (EMG) from three muscles (biceps femoris, gastrocnemius and tibialis anterior) were recorded during the exercise from six subjects. Integrated EMG (iEMG) and mean power frequency (MPF) were used to describe muscle activity. The T(re) was virtually unchanged during rest in every ensemble, whereas during exercise T(re) was significantly lower in SS than in LL. Mean skin temperature (T(sk)) decreased during rest in every ensemble, being significantly lower in SS than in LL. After the rest period local T(sk) of thigh and calf were significantly lower in SS than in LL and they were also lower in LSu than in LSc. At the beginning of the exercise the iEMG of the tibialis anterior muscle in SS and LL averaged 84 (SEM 7) and 64 (SEM 3) mu V (P <0.05), respectively. Respective values for LSu and LSc were 86 (SEM 9) and 66 (SEM 6) mu V (P <0.05). The MPF of the tibialis anterior muscle was significantly higher in LL 102 (SEM 5) Hz than in SS 90 (SEM 5) Hz (P <0.05) and similarly the MPF of the gastrocnemius muscle was also higher in LL 111 (SEM 5) Hz than in SS 100 (SEM 5) Hz (P < 0.05). It was concluded that exposing bare legs to a cool environment enhanced the motor unit activity in relation to covered legs. This would suggest that wearing shorts in a cool environment may, at the beginning of exercise, result in higher (about 25 percent) EMG activity and this may reflect increased muscle strain in comparison with wearing long trousers. Our results showing a unilateral increase in EMG activity during unilateral cooling suggest that the increase of strain is restricted to the uncovered part of the limb.