BACKGROUND This study was conducted to determine if the differences between efficacies of three treatments for immersion hypothermia are affected by body composition. METHODS Twelve subjects were divided into equally sized low (LF) and high (HF) fat groups. On three occasions subjects were each immersed in cold water until esophageal temperatures (Tes) decreased to approximately 33.2 degrees C (LF) and approximately 35.8 degrees C (HF). They were then rewarmed by: 1) shivering; 2) application of external heat; or 3) treadmill exercise in a balanced design. RESULTS For HF, the afterdrop during exercise (1.04 +/- 0.2 degrees C) was greater than during shivering (0.35 +/- 0.3 degrees C) and external heat (0.36 +/- 0.1 degree C) (p < 0.01). In LF, however, the exercise afterdrop (0.75 +/- 0.2 degree C) was greater than only external heat (0.35 +/- 0.2 degree C) (p < 0.05) but not shivering (0.58 +/- 0.4 degree C). There was a positive relationship between % fat and afterdrop for the exercise condition with a slope (95% C.I.) of 0.03 (0.01 to 0.05) degree C.% fat-1 (r2 = 0.37, p < 0.05). The exercise rewarming rate (3.48 +/- 1.1 degrees C.h-1) was greater (p < 0.01) than during both shivering (1.80 +/- 0.7 degrees C.h-1) and external heat (2.22 +/- 0.7 degrees C.h-1) in HF while no difference was seen between the three treatments (5.28 +/- 0.4, 4.86 +/- 1.1 and 5.16 +/- 0.7 degrees C.h-1, respectively) in LF. There were inverse relationships between % fat and rewarming rate in the exercise -0.12 (-0.23 to -0.01) degree C.h-1.% fat-1, (r2 = 0.38), shivering -0.27 (-0.38 to -0.16) degrees C.h-1.% fat-1, (r2 = 0.76) and external heat -0.26 (-0.35 to -0.17) degree C.h-1.% fat-1, (r2 = 0.83) conditions (p < 0.05). CONCLUSIONS The inter-treatment differences between these techniques are accentuated in the HF, and attenuated (afterdrop) or even eliminated (rewarming rate) in the LF subgroup.