Eight men and 8 women trained 3 days/ week for 16 weeks by doing attempted ballistic unilateral ankle dorsiflexions against resistance which either rendered the resultant contraction isometric (one limb) or allowed a relatively high velocity (joint angular velocity of 5.23 rad.s- 1 ) isokinetic concentric contraction (other limb). Training sessions consisted of 5 sets of 10 contractions of each type. Pre and post-training tests of peak torque at 0, 0.26, 0.52, 1.04, 1.55, 3.02, 4.19, 5.23 rad.s-1 indicated a velocity specific training response (p<0.01), with increases of -5.9, 5.6, 8.6, 15.3, 13.9, 14.1, 19.3, and 27.4% respectively. In a separate test, maximal voluntary isometric peak torque (6 .1%) and maximum rates of torque development (20.4%) and relaxation (31.5%) increased after training (p<0.01). Electrically evoked isometric tetanic peak torque and rate of torque relaxation did not change but rate of torque development increased 12. 6% ( p<0. 01). Evoked peak twitch torque did not change but time to peak torque and 1/2 relaxation time decreased 6.2 and 11.9% respectively (p<0.01). Electromyographic (EMG) recordings of the agonist tibialis anterior (TA) during test contractions showed no change in integrated EMG, but there was an increase (14.6%, p<0.05) in antagonist soleus (S) EMG from mid-test to post-test. The velocity specific response to the isometric and isokinetic concentric training modes was the same, indicating that it was the intent to make a ballistic contraction, rather than the resultant velocity of contraction, that determined the velocity specific response. The data also suggest that both neural and muscular adaptations contributed to the velocity specific training response / Thesis / Master of Science (MS)