For 30 days five healthy men aged 23-57 yr consumed a diet adequate in zinc (8.6 mg/day); they ate a low-zinc diet (3.6 mg/day) for the next 120 days and then received a zinc-supplemented (33.6 mg/day) diet for 30 days. Copper intake was constant at 1.8 mg/day. Aerobic capacity was determined periodically during each diet period. Relative zinc balance (% of control) declined during depletion (r = -0.28, P less than 0.009). Pre- and postexercise zinc concentrations decreased when dietary zinc was restricted (r = -0.61, P less than 0.0001 and r = -0.78, P less than 0.0001) and increased with supplementation (r = 0.61, P less than 0.008 and r = 0.76, P less than 0.0003, respectively). Both plasma zinc and hematocrit increased (P less than 0.01) after maximal exercise. To minimize the effect of hemoconcentration during exercise, the van Beaumont quotient (J. Appl. Physiol. 34: 102-106, 1973) was calculated using pre- and postexercise hematocrit and plasma zinc. The initial quotient of 1.8 +/- 1.8% (mean +/- SE) declined (P less than 0.05) to -7.4 +/- 2.3% during depletion. With zinc repletion, the quotient increased to 6.9 +/- 3.6%, which was greater (P less than 0.05) than the quotient in depletion but similar to the initial quotient. The quotient was a strong predictor (r = 0.71, P less than 0.0005) of the change in relative zinc balance during zinc depletion. In contrast, no changes were found in plasma copper content. These data suggest that zinc mobilization from tissues is impaired during zinc depletion, and they validate the use of the van Beaumont quotient as an index of change in body zinc stores.