The NCS defines series crimes as three or more incidents similar in nature in which the victim is unable to furnish details of each incident separately. Series crimes pose a reporting problem for victimization surveys and consequently are excluded from estimates based on NCS data. Persons who reported a series crime in their NCS interview between January and June 1985 were reinterviewed with a specially designed questionnaire. In about 60 percent of the series incidents, respondents recalled the details of each incident during the interview so that separate incident reports were obtained. In nearly half the cases, the number of incidents in the series between the original and followup interviews did not change. In over 75 percent of the series incidents, all fell into the same specific crime category. Violent series crimes that occurred in connection with a person's job were usually committed by different offenders who were generally strangers. In contrast, violent crimes committed by spouses or acquaintances were usually perpetrated by the same person who was well known to the victim. The report recommends raising the number of incidents needed to qualify as a series to six, improving NCS interviewers' training, and conducting additional research on the series crime reporting problem. 6 tables. (Author summary modified)