This study tests whether playing violent video games leads to desensitization and increased cardiovascular responding. In a laboratory experiment, 42 men spent 20 min playing either a high- or low-violence version of a “first-person shooter” game. Arousal (heart rate, respiration rate) was measured continuously. After playing the game, emotional responses to aversive and aggressive stimuli - pictures from Lang, Bradley, and Cuthbert’s (1999) International Affective Picture System - were assessed with self-ratings and physiological measurement (skin conductance). Results showed no differences in the judgments of emotional responses to the stimuli. However, different effects of game violence emerged in the physiological reactions to the different types of stimulus material. Participants in the high-violence condition showed significantly weaker reactions (desensitization) to aversive stimuli and reacted significantly more strongly (sensitization) to aggressive cues. No support was found for the arousal hypothesis. Post-hoc analyses are used to discuss possible moderating influences of gaming experience and player’s trait aggressiveness in terms of the General Aggression Model ( Anderson & Bushman, 2001 ) and the Downward Spiral Model ( Slater, Henry, Swaim, & Anderson, 2003 ).