Sir:I read with interest the recent article by Dore´et al. (2001), in particular the section that suggested: ''tissues other than solely lean leg muscle mass (such as the arms, trunk and gluteus maximus) contribute to high intensity cycling performance, and that the external resistive force should be matched to the capability of the active muscle tissue''.I would like to elaborate a little on the contribution of the upper body to high intensity leg power measured using cycle ergometry and the selection of resistive forces.We recently examined the performance of the legs when influenced by activity in the upper body, as shown by handgrip, during high intensity cycle ergometry, and the relationship between total body mass (TBM) and fat free mass (FFM). The with-grip protocol yielded significantly greater (P<0.05) peak mechanical power output than the without-grip protocol, suggesting a significant contribution from the upper body to the maximal power output measured in the legs (Baker et al. 2001b). In addition, as a first step towards quantifying the involvement of the upper body during leg cycle ergometry, surface electromyographs of the forearm musculature were recorded whilst performing each of the test protocols. During the with-grip ergometer tests, the intensity of the electrical activity recorded from the forearm musculature was greater than the intensity of electrical activity recorded from the forearm musculature during 100% isometric voluntary handgrip dynamometer contractions. This suggests maximal isometric-type forearm muscle contraction during the with-grip leg ergometry tests.Researchers should be mindful of this observation both when allocating ergometer loads, and when analysing blood borne metabolites.Isometric contraction has been shown to cause occlusion of the blood vessels passing through and between the activated muscles (Libonati et al. 1998). It is likely that blood samples taken from an antecubital vein, immediately after exercise, would also contain some blood from a tissue that had been occluded, and would not be representative of leg muscle blood. Equally, the subsequent reperfusion of arm musculature during relaxation would have an unknown effect on the concentration of blood-borne metabolites.We have also examined resistive force selection based on TBM and FFM (Baker et al. 2001a) using a traditional handgrip protocol. Increases (P<0.05) in mean (SD) peak power output (PPO) were found using the FFM protocol [1,015 (165) W TBM vs 1,099 (172) W FFM].Significant decreases (P<0.05) were observed for both resistive force selection [7.6 (1.4) kg TBM vs 6.7 (1.1) kg FFM] and for the time taken to reach PPO [3.8 (1.4) s TBM vs 2.9 (1) s FFM]. However, pedal velocity increased (P<0.01) when the two protocols were compared [129.4 (8.2) revolutionsAEmin -1 TBM vs 136.3 (8) revolutionsAEmin )1 FFM]. Rating of perceived exertion was (P<0.05) greater for FFM [18.4 (1.6) TBM vs 19.8 (0.4) FFM]. No differences in power or associated values were found between protocols for mean power output , fatigue...