This study was designed to find out if the lymph flow, indicated as albumin clearance, from active skeletal muscle is maintained constant during a prolonged steady-state exercise. 99mTc-labelled albumin was injected bilaterally into the vastus lateralis muscles of eight endurance-trained men. The radioactivity at the injection site was monitored by a gamma-camera before, frequently during, and after a 2 h run at a controlled steady intensity of 69 ± 4 % of the maximal heart rate. The fractional clearance rate of albumin was calculated for each monitoring interval, and was expressed as percentage clearance per minute (% min-1). During the first 15 min of exercise the clearance rate was five times higher than at rest before the exercise (0.29 ± 0.12 vs. 0.06 ± 0.05 % min-1, P = 0.001). During the next 25 min of running the clearance rate fell to 0.19 ± 0.08 % min-1 (P = 0.02), from which level it was further attenuated, being 0.12 ± 0.04 % min-1 at the end of the exercise. After the exercise the clearance rate fell rapidly to the level of 0.04 ± 0.03 % min-1. The results showed that the albumin clearance from working muscles is not constant during steady-state exercise, suggesting that lymph flow from exercising skeletal muscle may not be constant despite lymphatic pumping being assumed to be unchanged during the course of exercise (i.e. constant exercise intensity and running speed).