Interstitial K+ ([K+]i) was measured in human skeletal muscle by microdialysis during exhaustive leg exercise, with (AL) and without (L) previous intense arm exercise. In addition, the reproducibility of the [K+]i determinations was examined. Possible microdialysis-induced rupture of the sarcolemma was assessed by measurement of carnosine in the dialysate, because carnosine is only expected to be found intracellularly. Changes in [K+]i could be reproduced, when exhaustive leg exercise was performed on two different days, with a between-day difference of approximately 0.5 mM at rest and 1.5 mM at exhaustion. The time to exhaustion was shorter in AL than in L (2.7 +/- 0.3 vs. 4.0 +/- 0.3 min; P < 0.05). Furthermore, [K+]i was higher from 0 to 1.5 min of the intense leg exercise period in AL compared with L (9.2 +/- 0.7 vs. 6.4 +/- 0.9 mM; P < 0.001) and at exhaustion (11.9 +/- 0.5 vs. 10.3 +/- 0.6 mM; P < 0.05). The dialysate content of carnosine was elevated by exercise, but low-intensity exercise resulted in higher dialysate carnosine concentrations than subsequent intense exercise. Furthermore, no relationship was found between carnosine concentrations and [K+]i. Thus the present data suggest that microdialysis can be used to determine muscle [K+]i kinetics during intense exercise, when low-intensity exercise is performed before the intense exercise. The high [K+]i levels reached at exhaustion can be expected to cause fatigue, which is supported by the finding that a faster accumulation of interstitial K+, induced by prior arm exercise, was associated with a reduced time to fatigue.