Compared with the induction of heat acclimation (HA), studies investigating the decay and re-induction of HA (RA) are relatively sparse and have yielded conflicting results. Therefore, 16 semi-nude men were acclimated to dry-heat by undertaking an exercise protocol in a hot chamber (dry-bulb temperature 46.1 +/- 0.1 degrees C; relative humidity 17.9 +/- 0.1%) on 10 consecutive days (HA1-10) in winter UK. Thereafter, the subjects were divided into two groups and re-exposed to the work-in-heat tests after 12 and 26 days until RA was attained (RA(12), n = 8; RA(26), n = 8). The exercise protocol consisted of 60 min of treadmill walking (1.53 m s(-1)) at an incline individually set to induce a rectal temperature (T (re)) of approximately 38.5 degrees C during HA1 (equating to 45 +/- 4% peak oxygen uptake), followed by 10 min of rest and 40 min of further treadmill exercise, the intensity of which was increased across HA to maintain T(re )at approximately 38.5 degrees C. T(re), mean skin temperature, heart rate and rate of total water loss measured at 60 min did not change after HA7, and HA was taken as the mean of the responses during HA8-10. For both groups, there was no decay in T(re) and for all measured variables RA was attained after 2 and 4 days in RA(12) and RA(26), respectively. It is concluded that once adaptation to heat has been attained, the time that individuals may spend in cooler conditions before returning to a hot environment could be as long as one month, without the need for extensive re-adaptation to heat.