Children ages 7 to 15 and adults were asked to detect changes in an unaccompanied drumbeat that either increased, decreased, or maintained the same tempo across measures. Marked differences in performance as a function of age were found for the change trials, older subjects being more accurate, while musical training was relatively unimportant. For trials in which the beat remained constant, however, subjects with musical training were more accurate, while age, by itself, had little effect upon performance. A multiprocess model for the perception of tempo is discussed.Tempo is a fundamental aspect of music, yet relatively few researchers have considered the development of sensitivity to differences in tempo. The existing literature suggests a complex picture. Some sensitivity to fluctuations in tempo can be found in the movements of very young children (Moog, 1976), although the extent of this tempo matching is quite limited. One might expect gradually increasing accuracy in the perception of tempo throughout the school years, consistent with a general picture of increase in musical sophistication with age. Existing research, however, does not suggest such a straightforward development.Wang and Salzberg (1984) administered a tempo discrimination task to string students ages 7 to 18. Tempo in musical excerpts increased, decreased, or remained constant, and listeners were asked to note both the point and direction of tempo change. Somewhat surprising, a curvilinear relationship was found between accuracy and age as well as between accuracy and years of musical training. Performance improved until age 13, after which it decreased. The mean number of correct responses also increased with years of musical training until approximately 5.5 years of training, after which it declined. The authors speculate that this curvilinear relationship may have been due to the small number of subjects older than 14 years and to the large amount of variability found in the scores of the older subjects. Also, as Wang and Salzberg acknowledge, age and musical experience were confounded in this study, and it was not possible to decide which of the two factors was primarily responsible for the results.