The Johnstone-El-Banna model of problem solving is based on workingmemory theory as well as on Pascual-Leone's M-space theory. The operation and validity of the model depends on a number of necessary conditions, such as a simple logical structure, availability and accessibility of the partial steps, absence of "noise," and lack of familiarity with the problem type. If these, and some other conditions, are not fulfilled, the model may not operate; that is, solvers may be successful, even if the informationprocessing demand (Z-demand) is greater than their information-processing capacity and, vice versa. Sixteen organic chemical-synthesis problems, with a simple logical structure and varying Z-demand from 2 to 8, were used in this work. We studied two samples of students (age 17-18), one without (N ϭ 128) and the other with (N ϭ 191) some previous training (at least in part) in organic-synthesis problems. Although the predicted pattern was observed in both samples, it was found that the model was more useful in the case of the students without previous training. Finally, the model predicts better with the fieldindependent and the field-intermediate students, but less so with the field-dependent ones.