To obtain a complete picture of recurrence patterns, families identified in an earlier research project funded by the National Center on Child Abuse and Neglect were followed for 5 years from their first 1988 report confirmed by a public child protective service (CPS) agency. Recurrence was defined as any confirmed report of physical abuse, sexual abuse, or neglect of any child in the family that occurred at least 1 day following the index incident report date. Sampling involved two phases: (1) sampling from the population of substantiated CPS reports in 1988 following a set of inclusion and exclusion criteria; and (2) applying a second set of exclusionary criteria to a select group of CPS families to follow with respect to recurrences. Results of survival analysis indicated that risk of recurrence was greatest during the 30 days following a report, that risk declined with time and remained relatively low for 2 years following the termination of services, and that recurrence depended on the type of maltreatment. Most recurrence families experienced only one recurrence, but a clustering effect was apparent for families that experienced multiple recurrences. 55 notes, 6 tables, and 5 figures