The effect of varying the body surface area being cooled by a liquid microclimate system was evaluated during exercise heat-stress conditions. Six male subjects performed a total of six exercise (O2 uptake = 1.2 l/min) tests in a hot environment (ambient temperature = 38 degrees C, relative humidity = 30%) while dressed in clothing having low moisture permeability and high insulation. Each subject completed two upper body exercise (U; arm crank) tests: 1) with only the torso surface (T) cooled; and 2) with the surfaces of both the torso and upper arms (TA) cooled [coolant temperature at the inlet (Ti) was 20 degrees C for all upper body tests]. Each subject also completed four lower body exercise (L; walking) tests: 1) with only the T cooled (Ti = 20 degrees C); 2) with only the T cooled (Ti = 26 degrees C); 3) with torso, upper arm, and thigh surface (TAT) cooled (Ti = 20 degrees C); and 4) with TAT cooled (Ti = 26 degrees C). During U exercise, TA cooling had no effects compared with cooling only T. During L exercise, sweat rates, heart rates, and rectal temperature (Tre) changes were less with TAT cooling compared with cooling only the T. Altering Ti had no effect on Tre changes, but higher heart rates were observed with 26 than with 20 degrees C. These data indicate that cooling arms during upper body exercise provides no thermoregulatory advantage, although cooling the thigh surfaces during lower body exercise does provide an advantage.