To clarify the origin of local cold adaptation and to define precisely its influence on muscle bio-energetics during local exercise, five subjects were subjected to repeated 5 degrees C cold water immersion of the right hand and forearm. The first aim of our investigation was therefore carried out by measuring local skin temperatures and peripheral blood flow during a cold hand test (5 degrees C, 5 min) followed by a 10-min recovery period. The 31P by nuclear magnetic resonance (31PNMR) muscle bio-energetic changes, indicating possible heat production changes, were measured during the recovery period. The second aim of our investigation was carried out by measuring 31PNMR muscle bioenergetics during handgrip exercise (10% of the maximal voluntary contraction for 5 min followed by a 10-min recovery period) performed both at a comfortable ambient temperature (22 degrees C; E) and after a cold hand test (EC), before and after local cold adaptation. Local cold adaptation, confirmed by warmer skin temperatures of the extremities (+30%, P less than 0.05), was related more to an increased peripheral blood flow, as shown by the smaller decrease in systolic peak [-245 (SEM 30) Hz vs -382 (SEM 95) Hz, P less than 0.05] than to a change in local heat production, because muscle bioenergetics did not vary. Acute local cold immersion decreased the inorganic phosphate:phosphocreatine (PC) ratio during EC compared to E [+0.006 (SEM 0.010) vs +0.078 (SEM 0.002) before acclimation and +0.029 (SEM 0.002) vs +0.090 (SEM 0.002) after acclimation respectively, P less than 0.05] without significant change in the PC:beta-adenosine triphosphate ratio and pH. Local adaptation did not modify these results statistically. The recovery of PC during E increased after acclimation [9.0 (SEM 0.2) min vs 3.0 (SEM 0.4) min, P less than 0.05]. These results suggested that local cold adaptation is related more to peripheral blood flow changes than to increased metabolic heat production in the muscle.