The aim of this research was to investigate the effects of different quantities of random and blocked varied practice, after constant practice, on adaptive process in motor sldll acquisition. Three experimenta were conducted manipulating the task component in the random and blocked varied practices: visual stimulus speed (experiment 1); sequential response pattem (experiment 2); and visual stimulus speed and sequential response pattem simultaneously (experiment 3). In ali experiments the design involved three quantities of varied practice (18, 36, and 63 trials), and two types of varied practice (random and blocked). The design also involved two phases: stabilization and adaptation. Participants were children, which performed a task that consisted of touching response keys sequentially in conjunction with a visual stimulus. The dependent variables were the absolute, variable, and constant error related to the task goal, and the relative timing and response time to evaluate response pattem macrostructure and micro-structure, respectively. Altogether, the results of the three experiments allowed concluding that: (a) there is an optimal levei of varied practice after constant practice. (b) This levei corresponds to eighteen random trials of three visual stimulus speeds or sequential response pattems, and to eighteen random and blocked trials involving the combination of three visual stimulus speeds and three sequential response pattems. (c) The optimal levei is dependent of the varied practice type (blocked or random) only if this refers to the manipulation of perceptive and motor task aspects separately.