Over the past ten years, sociologists have broadened their view of what counts as an appropriate measure of mental health. This reflects a growing recognition that individuals express emotional upset in various ways. For example, some individuals are more likely to become depressed in response to stress while others are more likely to drink heavily. Contemporary studies often include measures of "internalizing" (i.e., more feeling-based measures) and "externalizing" (i.e., more behavioral measures) styles of psychopathology, especially when studying group difference in mental health. Alcohol abuse is the classic measure of externalized distress in sociological research. In this paper, we present a theoretical argument and supporting empirical evidence to argue that violent behavior should be included as a measure of externalized distress in response to stress.