Previous studies have shown that inhibiting negative or positive emotion-expressive behavior leads to increased sympathetic activation. Inhibiting facial behavior while in an affectively neutral state has no such physiological consequences. This suggests that there may be something special about inhibiting emotion-expressive behavior. To test the boundary conditions of the suppression effect, acoustic startles were delivered to 252 participants in three experimental groups. Participants in one group received unanticipated startles. Participants in the other two groups were told that after a 20-s countdown a loud noise would occur; participants in one of these groups were further told to inhibit their expressive behavior. Results indicated that startle suppression increased sympathetic activation. These findings extend prior work on emotion suppression, and suggest that inhibiting other biologically based responses also may be physiologically taxing.