The study used data collected in the Charlotte Spouse Assault Replication Project (SARP) in North Carolina. Like the other SARP experiments, the Charlotte experiment focused on the misdemeanor range of cases in which police were authorized, but not required, to make warrantless arrests. Included in the study were heterosexual couples who had a current or previous marital or cohabiting relationship. All of the 424 cases sampled involved heterosexual couples with female victims and male offenders. The study examined whether characteristics of the offenses, such as cause of the argument and victim injury; offender characteristics, such as prior record and substance abuse; or victim characteristics, such as relationship with the offender and substance abuse, most impact the prosecution's decision to prosecute. The two significant determinants of the decision to prosecute were legal factors strongly associated with the probability of achieving a successful prosecutorial outcome. These two factors were victim injury and a strong victim preference for no official criminal justice intervention in the case. Victim injury provides clear corroborative evidence of the victim's allegation that she has been harmed by the defendant and is likely to facilitate the probability of the defendant's conviction. On the other hand, the victim's resistance to cooperating in the prosecution of the abusive spouse has a negative impact on the probability of conviction. 3 tables and 35 references