The purpose of this study was to investigate the thermoregulatory mechanisms underlying artificial acclimatization to cold and to compare them with those of naturally acclimatized men. Six white men were cooled, nude, in air at 10 degrees C for 2 h before and after they had been acclimatized by ten daily cold (15 degrees C) baths of 30-60 min followed by rapid rewarming in hot (38-42 degrees C) water, and again 4 months later after acclimatization had decayed. Six control subjects also underwent the same tests, providing an opportunity to discriminate between changes caused by the immersions and those caused by extraneous influences. Acclimatization significantly reduced heat production and heat loss (P < 0.05) but did not change heat debt. The reduced heat production was accompanied by reductions in shivering (P < 0.10) and in cold-induced muscle tenseness; no evidence of nonshivering thermogenesis or active brown fat was found. These findings are attributed to increased tissue insulation, mediated by an enhanced vascular response to cold that did not involve the cutaneous circulation and was probably located in skeletal muscle. Thermal sensation and discomfort did not change, although perceived strain tended to increase (P = 0.08). Acclimatization was accompanied by, but was unrelated to, slower cooling of the finger and toe. The main conclusions, and many specific findings, agree with those of two previous studies made by the same techniques in naturally acclimatized men wintering in Antarctica. Other significant findings included changes--in particular reduced thermoneutral rectal temperature and a delayed onset of shivering--that are commonly regarded as evidence of acclimatization but were in fact unrelated to it as they also occurred in the control group. They are attributed to extraneous influences, in particular the relaxation of heightened arousal ('first-time effects') found in the baseline tests.