Sixteen male students exercised for 14 days (1 h/ day) in the heat for heat acclimation (HA). During de-acclimation (DA) one group exercised in the cold (EXG, n = 8) for 60 min/day (morning) and was exposed to the cold for another hour (afternoon) for 14 days. The other group was exposed to the cold (EPG, n = 8) for 1 h each in the morning and afternoon (Ta: 18.0 degrees C, RH: 58%) over the same period. All returned to exercise in the heat for reacclimation (RA) for 10 days. Subjects were tested on days 1, 16 21, 32, 36 and 44 on a bicycle ergometer for 60 min at 60% of VO2max in the heat (Ta: 31.1 degrees C, RH: 70%). Rectal temperature (T-re) and heart rate (HR) a 40 min of exercise were used to determine the decay/gain of HA, which was calculated using the formula described by Pandolf et al. (Ergonomics, 20:399-408, 1977). After HA (day 16) T-re and HR decreased significantly. During DA, EXG showed decay in T-re of 24 and 35% and HR of 29 and 35% on days 21 and 32, respectively. For EPG the corresponding decay was of 2 and 9% for T-re and 17 and 17% for HR. After 10 days of RA, EXG showed gains of 11% in T-re and 12% in HR, while EPG showed gains of 47% in T-re and 38% in HR. In conclusion, EXG had greater decay during DA and lower gains in RA coin pared to EPG. However, the differences between groups were significant only for T-re after 4 days of DA.