Data were obtained from a questionnaire survey and official police statistics. Of 730 survey packages distributed, 379 were returned for a response rate of 52 percent. Both cross-tabulation and regression methods were used to analyze survey data. Findings indicated that police officers were more likely to arrest when the victim accused the suspect of assault, the victim pressed charges against the suspect, the suspect employed a weapon, and the suspect was hostile to police officers. Alcohol consumption by either suspect or victim had no significant effect on the police officer's decision to arrest. Further, the police officer's decision to arrest was not influenced by whether the victim was married to the suspect or if the victim called the police. Police officers were most likely to arrest in compelling cases and least likely to arrest in trivial cases. The likelihood of police arrest increased if the suspect was accused of assault and had a weapon and if the suspect was accused of assault and was hostile to police officers. It was determined that police discretion in domestic disturbances is influenced by legal and extralegal variables and that police officers base their decision to arrest mainly on legal criteria. A detailed review of the literature on police intervention in domestic disturbances is included. 81 references, 23 tables, and 6 figures