Computerized agility training (CAT) products are frequently suggested to improve agility. However, these claims often are made without unbiased scientific support. Therefore, the purpose of this study was (a) to determine the reliability and effectiveness of a 4-week CAT training program on foot speed (FS) and choice reaction (REACT), and (b) to assess whether training on the CAT would facilitate the improvement of a separate change-of-direction (COD) test in non-agility-trained, but active, men and women. Twenty-three participants (15 men, 8 women) pre- and posttested on FS, REACT, and COD drills. Eleven of those participants (7 men, 4 women) engaged in 4 weeks of training on the FS and REACT drills (EX). The remaining 12 (8 men, 4 women) did not participate in the training and served as controls (CON). Coefficients of variation indicate strong precision for FS (6.9%) and REACT (2.6%). Test-retest reliability, as analyzed by intraclass correlations (ICC), were high for both FS and REACT (0.89). Significant test-by-group interactions were observed for all three tests: FS (p = 0.004), REACT (p = 0.011), and COD (p = 0.049). Post hoc analysis indicated that EX increased foot contacts for the FS drill (p = 0.006), whereas REACT and COD demonstrated decreases in time to completion (p = 0.013 and 0.038, respectively). The CON group did not improve on any of the tests. This study indicates that the chosen CAT is an accurate and reliable tool for measuring foot speed and reaction time. These data justify the use of this CAT in analyzing foot speed and reaction time. Altogether, 4 weeks of foot speed and reaction training on the chosen CAT produced improvements in overall agility in non-agility-trained, but active, men and women. These data warrant the integration of such a device into the training program of untrained athletes attempting to improve agility.