The decision to make a drawing, we think, depends primarily on the assessment that doing so will make the solution easier to find. The perceived benefits might be that a drawing mitigates working memory, presents a concrete model on which to act, eases the search for related information, or makes problem characteristics more explicit. Those benefits do not apply (or occur) to young children trying to solve simple word problems. For fifth graders, the benefits of making drawings of the kind of problems we presented were much more apparent. How these benefits are perceived, however, depends on the problem. A drawing of a quantified three-term series problem, for instance, relieves working memory and visualizes the unknown relationship that needs to be quantified. The benefit of a drawing of an (a-1).b problem is that it serves to make explicit that the number of intervals between adjacent objects is one less than the number of objects. In the case of fifth graders, the utility of a drawing also is more readily perceived because many of the concepts in complex word problems have been taught with the help of drawings (e.g., area, volume, and fractions). This cues the making of a drawing in the solution process, especially when nonroutine problems are concerned. In all those cases, however, the decision to make a drawing depends on the benefits perceived by the pupil. We sometimes observed that pupils decided not to make a drawing because they had immediate access to a solution procedure, even though this procedure was incorrect (and could be corrected when a drawing was made). This situation often occurred when pupils were confronted with an (a-1).b problem. The belief in the correctness of the chosen solution procedure sometimes was so strong that, when asked to make a drawing of the situation, the drawing was adapted to fit the solution procedure.