Attempts to suppress stereotypes have often been found to result in an increased accessibility of these stereotypes. According to thought suppression literature together with research on prime-to-behavior effects, we hypothesized that suppression of stereotype can lead people to subsequently behave in accordance with its content and that these effects are stronger after suppression (rebound) than after a classical priming condition (i.e., no-suppression condition). Experiment 1 showed that suppression of the stereotype of sportsmen (associated with poor math performance) but not of Italian men (not related to math performance) led participants to subsequently perform worse on a calculus task in comparison to non-suppressors. These effects were replicated in a second experiment with another stereotype (elderly) and another behavior that does not require self-regulation (walking speed): Suppressors walked slower than non-suppressors. These findings are considered in the context of mental control and social stereotyping. Copyright © 2009 John Wiley & Sons, Ltd.