1. Hand-warming is employed during metabolic studies to "arterialize' venous blood, but also has systemic haemodynamic effects. Study 1 aimed to determine if hand-warming affects the value for whole-body insulin sensitivity derived from the hyperinsulinaemic euglycaemic clamp technique. Study 2 was designed to assess the effect of hand-warming on contralateral forearm blood flow. 2. In study 1, eight healthy male subjects attended for four modified euglycaemic clamp studies, during which the right hand was placed in a heated-air hand box (55 degrees C), and the glucose infusion rate was adjusted according to blood samples from a cannula in either an ipsilateral dorsal hand vein or a contralateral antecubital vein with the box switched either on or off. In study 2, five healthy subjects attended two study days when the effect of 2 h of hand-warming (or control) on contralateral forewarm blood flow was measured. 3. Study 1; when clamps were performed according to samples taken from the contralateral antecubital vein, insulin sensitivity values were significantly lower during box-on versus box-off clamps (mean +/- SD 10.2 +/- 3.0 versus 13.0 +/- 3.8 mg min-1 kg-1, P < 0.05, 95% confidence interval -0.2, -5.3). When clamps were performed according to samples taken from the ipsilateral hand, there was no difference in insulin sensitivity during box-on versus box-off clamps (9.2 +/- 2.1 versus 9.0 +/- 1.7 mg min-1 kg-1, P not significant, 95% confidence interval -0.5, +0.9). There was a mean increase in heart rate of 6 beats/min in the box-on conditions (P < 0.05, analysis of variance). Study 2; forearm blood flow in the contralateral arm during hand-warming was significantly higher than in the control condition (P < 0.05, analysis of variance), although heart rate was similar on both study days. 4. Hand-warming had a detectable effect on insulin sensitivity when clamps were performed according to samples withdrawn from the contralateral arm, but no measurable effect when clamps were performed in the conventional manner. In addition, hand-warming increased heart rate during hyperinsulinaemia and contralateral FBF under basal conditions. These findings raise concern about unwanted (potentially confounding) systemic haemodynamic effects of hand-warming on the measurement of insulin sensitivity and insulin-mediated vasodilation.