This study aimed to examine the reliability and sex- and age-related differences of step tests with stipulated tempos as well as to clarify useful test parameters and tempos. One hundred forty elderly people and fifty young adults conducted tapping and stepping tests, matching the tempo provided by a metronome. Both tests involve movements where the subject touches a sheet with both the right and left hands or right and left legs at a designated spot. Evaluation parameters were the time difference between the beep sound and the time at which the sheet was touched in both tests as well as two-leg support and one-leg support times in the step test. The trial-to-trial reliability of the parameters in both tests was high. The time differences of both 40 bpm tests in the elderly were larger in males than in females. In the step test, the time difference and two-leg support times of the elderly were larger, in the order of 40, 60, and 120 bpm, and the one-leg support time was less in 40 bpm than 60 bpm or 120 bpm. The one-leg support time of the young subjects was larger, in the order of 40, 60, and 120 bpm. A significant age-related difference was found in the 40 bpm and 60 bpm test, and the time difference and two-leg support time were larger in the elderly while the one-leg support time was larger in the young subjects. The time difference at 40 bpm in the elderly was larger in the step test than in the tap test. There was no significant difference between both tests in the young subjects. In conclusion, the step test with the slow tempo, because it requires a long one-leg support phase, is effective for evaluating dynamic balance in the elderly. The time difference and two-leg and one-leg support times are effective evaluation parameters of the step test.