Three scales were assessed as indices of the frequency, prevalence, and severity of violence. There appears to be a strong relationship between severity of violence inside the home and involvement in extradomestic violence, but this factor was not useful in explaining violent behavior. Among background variables, exposure to violence during childhood was a strong and consistent predictor of both severity and prevalence. Age is both a strong predictor and a suppressor variable, with younger men at risk for more severe and widespread violence. Education predicts severity of victim injury, particularly for younger batterers. While there are indications that whites are more often violent, this may result from a sampling artifact. Unemployment, which was high in the sample, was associated with severity of victim injury. Frequency of violence, frequency of injury, and abuse during pregnancy are predictive of both prevalence and severity of abuse. The data also suggest that frequency of violence in the home is predictive of the incidence of violence outside the home. Actual violence appears to be a strong predictor of subsequent violence; however, the data also suggest a relationship between the frequency and severity of violence and its prevalence. Policy implications of the findings are briefly discussed. Tabular data, 31 references, and 9 footnotes are provided.