A major content analysis study of violence on television in Britain during 1994/95 explored different quantitative measures for expressing the amount of violence on different channels. Programs were video-recorded from eight TV channels, four terrestrial (BBCI, BBC2, ITV, Channel 4) and four satellite channels (Sky One, UK Gold, Sky Movies, The Movie Channel) over four separate weeks between early October 1994 and mid-February 1995. A total of 4,606 programs and 4,715 hours of TV output were coded, using an a priori definition of violence. The amount of violence on TV was quantified according to several distinct types of measure, including the percentage of programs which contained any violence, the total number of violent incidents, the average number of violent incidents per violence-containing program, the average rate of violent incidents per hour, the amount of time occupied by violence, the proportion of total program-running time occupied by violence, and the extent to which certain proportions of violence were accounted for by particular programs. Results showed that these different measures could give quite distinct impressions about how much violence occurs on TV. Certain popularly cited measures, such as percentages of programs with any violence and average rates per hour may give a misleading impression of the prevalence of violence on TV. Other less often cited measures, such as violence minutage as a proportion of total running time of broadcast output, and the disproportionate contribution to overall violence levels of small numbers of selected programs place the issue in perspective and reveal that violence occurs heavily in certain parts of TV output, while being almost non-existent over large sections of the mainstream schedules.