Forearm-fingertip skin temperature differentials (T(sk-diff)) are used to indicate vasomotor tone, vasoconstriction defined as having occurred when T(sk-diff)> or =4 degrees C (Sessler et al. 1987, 1988a, b). This study was conducted to determine whether T(sk-diff) or finger pad heat flux (HF) can be used to predict when vasoconstriction and vasodilatation occur. Seven subjects (one female) sat in water at [mean (SD)] 40.7 (0.8) degrees C until their core temperature (T(c)) increased by 1 degrees C, ensuring vasodilatation. The water was then cooled [at a rate of 0.6 (0.1) degrees C x min(-1)] until T(c) fell to 0.5 degrees C below pretesting values, causing vasoconstriction. Subjects were then rewarmed in water [41.2 (1.0) degrees C]. Skin blood flow (SkBF) was measured using laser Doppler flowmetry (LDF) on the left second finger pad [immersed in water at 10.4 (1.4) degrees C as part of another experiment], and infrared plethysmography on the third finger pad of both hands. T(sk-diff) and HF were measured on the right upper limb, which remained in air. When vasodilated, the subjects had a stable T(sk-diff) and HF. During cooling, rapid-onset vasoconstriction occurred coincidental with large gradient changes in HF and T(sk-diff) (inflection points). In two subjects the original vasoconstriction definition (T(sk-diff)> or =4 degrees C) was not attained, in the other five this was achieved 31-51 min after vasoconstriction. During rewarming, the T(sk-diff) and HF inflection points less accurately reflected the onset of vasodilatation, although with one exception they were within 5 min of the LDF changes. We conclude that T(sk-diff) and HF inflection points predict vasoconstriction accurately, and better than T(sk-diff)> or =4 degrees C.