This study reconsiders a series of drawing tasks (Goodnow, 1978) in which children have to modify their stereotypical drawing of the human figure to represent a person in movement. Another task, in which children have to differentiate the drawing of a kangaroo from that of a person, is also considered. According to a neo-Piagetian model of drawing development (Morra, 1995), it is hypothesized that three factors underlie children's ability to flexibly modify their drawings: (a) the amount of attentional resources (M capacity) that a child can use to activate task-relevant figurative and operative schemes; (b) automatic activation of figurative schemes from perceptual input that can be obtained by presenting a model; and (c) activation of executive schemes that set appropriate goals and monitor performance, which can be obtained by manipulating contextual variables, such as task order. Three experiments (with a total of 645 participants in the age range from 5 to 9 years) tested successfully the theoretical predictions about the causal factors of drawing flexibility.