The increase in motor unit force that occurs with aging has been hypothesized to cause a decline in the ability to maintain a constant submaximal force. To test this hypothesis, young and elderly subjects performed a 12-wk strength-training program that was intended to increase motor unit force. The training program caused similar increases (%initial) in the training load (137.4 +/- 17.2%), twitch force (23.1 +/- 7.4%), and maximum voluntary contraction force (39.2 +/- 6.8%) of the first dorsal interosseus muscle for the young and elderly subjects. The increase in strength was associated with a modest increase in muscle volume (7% of initial value) and a nonmonotonic increase in the surface-recorded electromyogram that was significant at week 8 but not at week 12. The elderly subjects reduced the variability in force at the lower target forces (2.5, 5.0, and 20.0% maximum voluntary contraction force). This improvement, however, was unrelated to changes in the distribution of motor unit forces, which was not consistent with the hypothesis that the greater coefficient of variation for the force fluctuations is due to increased motor unit forces.