A study was conducted in which half-hour segments of prime-time network dramatic programing were tabulated for such violence-related items as: the seriousness and significance of violence, the number of violent actions in the program, and the duration of violence. Other factors considered were: audience size, share of audience, program duration, program type, program tone, format, and time of broadcast. This consideration of how one dimension of program content (violence) affects audience size indicated that violence measures accounted for little more than 5% of the variance in audience shares, an additional 4% during nonfamily hour programs, and 55% when the focus was only on feature films. (TJ)