Seven healthy male subjects performed 10 maximal 6-s sprints, separated by 30-s recovery periods, on a non-motorized treadmill. On two occasions, separated by 3 days, the subjects ingested a solution of either sodium bicarbonate (NaHCO3; alkaline) or sodium chloride (NaCl; placebo), 2.5 h prior to exercise. The doses were 0.3 g kg-1 body mass for the alkaline treatment and 1.5 g total for the placebo, dissolved in 500 ml of water. The order of testing was randomly assigned. Pre-exercise blood pH was 7.43 +/- 0.02 and 7.38 +/- 0.01 for the alkaline and placebo trials respectively (P less than 0.01). Performance indices (i.e. mean and peak power outputs and mean and peak running speeds) were significantly reduced as a result of the cumulative effects of successive sprints, but not significantly affected by the treatments. However, the total work done (i.e. mean power output) in the alkaline condition was 2% higher than that achieved in the placebo condition. Post-exercise blood lactate concentrations were higher for the alkaline treatment than for the placebo condition (15.3 +/- 3.7 vs 13.6 +/- 3.0 mM respectively; P less than 0.01), but blood pH was similar in both conditions (alkaline: 7.15 +/- 0.13; placebo: 7.09 +/- 0.11). In both conditions, a relationship was found between post-exercise blood lactate and mean power output (alkaline: r = 0.82, P less than 0.01; placebo: r = 0.79, P less than 0.01). No significant differences were found in VE, VO2 and VCO2 between the two experimental conditions. This study demonstrates that alkali ingestion results in significant shifts in the acid-base balance of the blood, but has no effect on the power output during repeated bouts of brief maximal exercise.