In 2 experiments female Ss observed a confederate solve 6 “water-jar” problems. Each problem permitted 2 modes of solution. In Exp. I the 3 conditions were observation of direct solutions, indirect solutions, or a mixture of both. In subsequent performances on similar problems, observers adopted the solution methods that had been observed even when inefficient. In Exp. II the order of the jar capacities was varied to test for the acquisition by observation of a more general solution set. The 3 conditions were observation of indirect solutions, direct solutions, or no observation (control). The results for both method of solution adopted and time for solution indicate that observation of the direct solution did not facilitate problem solving, but observation of the indirect solution did hinder this activity.