This study examined problem-solving abilities across preschool-age children using the Tower of London (TOL) task. The TOL is a spatial puzzle, which requires planning, an executive function, along with the ability to adhere to a set of rules to solve it successfully. Sixteen TOL problems of various difficulty levels were administered to 25 preschool children, split into young (ages 3.0-4.4, n = 12) and old (ages 4.5-5.10, n = 13) age groups, thereby generating an Age (young vs. old) Yen Difficulty Level (4 difficulty levels) experimental design. The following dependent measures were coded from videotape: latency until the 1st move, total time taken to solve, percentage of rule violations, total number of moves made, and percentage solved correctly. There were significant main effects of difficulty level for all dependent measures. In addition, an Age Yen Difficulty Level interaction, F(3, 22) = 4.15, p < .01, suggested that the older group's greater speed and accuracy was especially prominent on the most difficult problems. These data imply that the ability to use some executive functions undergoes significant development during the preschool years and is a function of both age and problem complexity.