In this study three issues related to "expert" problem solving were studied; the problems considered were based on Tower of Hanoi puzzles. A computer simulation was proposed and tested, effects of cues on subjects' performance were studied, and inter-problem transfer was analyzed. Four predictions were posed for the simulation study. Two of these were upheld, relating problem difficulty to amount of memory needed and number of subgoals necessary. The others concerned "start-up tasks" and recognition of previous stages; results concerning these led to a modification of the simulation. In the second phase of the study, subjects were given three different sets of instructions for problem solving (ambiguous, subgoals, sequence). There was no significant difference in the performance of the three groups. After some experience with the problems, these groups were compared with an inexperienced group for the transfer study. Results favored the view that solution planning was learned during practice and applied on the transfer task. (Author/SD)