Kirschner, Sweller and Clark (2006) suggest that minimally guided instructional methods are not as successful or efficient as guided instructional approaches. The goal of the current study was to address both the issue of learner performance and knowledge transfer when problem solving in anchored instruction was used for mathematical instruction as compared to using worked examples in an anchored instruction context. The participants in the study were 50 fifth grade students at an inner-city, high-achieving, Title I school. The participants were randomly assigned to groups that either studied the solution to the problem presented in two different series of The Adventures of Jasper Woodbury, or solved the problems using problem-solving strategies following the anchored instruction approach. There were two measures for each of the two problems. First, the participants were asked to recall the solution to the problems that they solved. Second, the participants were asked to solve transfer problems that had the same basic procedures as the given problem. Independent t tests were carried out to examine the differences between the two groups on both recall problems and both transfer problems. Significant differences were found for learning and transfer in favor of the problem-solving group over the worked example group. The current study is significant in that it expands the knowledge base in the area of instructional techniques and strategies. The current study supports the use of anchored instruction in the area of mathematics, and suggests that problem processes are not superfluous, and studying a solution may not provide as much opportunity for learning and transfer.