The overarching study hypothesis is that the greater the ease of transition into the new noncustodial parental role, the less the probability of parental kidnapping. Subjects were the entire population of married couples with children under age 12 who filed for divorce in Hennepin County, Minn., in 1979. Of the 2,092 questionnaires mailed to custodial parents, 503 completed questionnaires were returned, out of which 61 incidents of parental kidnapping were reported. A more detailed questionnaire was then sent to the custodial parents who had experienced a kidnapping. Questionnaires were also sent to 56 noncustodial kidnapping parents and a sample of 108 noncustodial, nonkidnapping parents. Only 14 questionnaires were returned from the kidnapping parents. Questionnaires solicited marital and family information, visitation history, support history, and the husband's work history. Information was also obtained on the kidnapped children. More than half the children had been previously kidnapped, and the majority of the children were returned by the kidnapping parents within 18 hours. Neither children's sex, age, or race were important predictors of parental kidnapping. The most significant factors in parental kidnapping were a contested divorce and whether child support was regularly paid. The role-transition hypothesis was not supported. Some alternative theories are offered to explain the findings. 3 figures, 85 tables, study instruments, and 300-item bibliography.