This paper explored the effects of different feedback manipulations on letter-string problem solving by analogical transfer. Specifically, the present studies assessed the effects of the type of feedback (first experiment); long exposure to feedback (second experiment) and the number of feedback problems during the practice phase (experiment 3). Participants were college undergraduate students who voluntarily received and answered printed letter string problems. Results showed that correct answer feedback has a significant positive effect on deep structure problem solving; they also showed that problem solution occurs much faster and evolves in a more straightforward manner with this type of feedback. Results also showed a direct relationship between the number of problems with correct answer feedback during the training phase, and the number of correct answers during the testing trial. Results are discussed in terms of the importance of using within-subject experimental designs in both the feedback interventions and analogical transfer literatures.