Each 1979 incident report was studied. Of the 638 incidents recorded, 182 met the criteria for assault, and 221 met the criteria for battery. Data were collected on the time and location of the incidents, preludes to the incidents, the conflicting parties, the weapon used, responses to the assailants, impact and injury, and offender characteristics. Data were used to compare the explanatory value of several competing models of the relationship between assaults and batteries; stages of interpersonal violence, degrees of interpersonal violence, individual patterns of interpersonal violence, and situational patterns of interpersonal violence. The four models are not mutually exclusive, and each is consistent with at least some of the study data. Although no final choice can be made between competing interpretations of the incidents, the observed differences between assaults and batteries reflect more than differences in the definitions used, the stage at which intervention occurred, the degree of violence involved, or the individual characteristics of the assailants. Some of the differences between assaults and batteries reflect differences in the situations to which the assailants and others were responding. Implications are drawn for the study of criminal offenses, the practice of combining attempted and completed offenses in crime classifications, and the interpretation of offenders' histories. 6 tables and 16 footnotes.