Much of the initial and continuing support for the selective incapacitation of chronic offenders derives from the study of the Philadelphia birth cohort, which found that 6 percent of the cohort accounted for more than half of the entire group's police contacts. This finding has been used as an argument for selectively imprisoning such chronic offenders. The problem in implementing such a policy is identifying the potential chronic offender. This paper reports on a study of the arrest records of the Philadelphia cohort of 9,945 males born in 1945 who lived in Philadelphia between the ages of 10 and 18 years. The focus is on the power of the arrest record in itself to predict future arrest rates. A prediction equation was developed and tested using regression analysis. The findings indicate that at the point where an existing arrest record is predictive of future arrest record, the highest-rate future offenders are only 2.4 times more criminally active than those projected to be the least active. Thus, although the findings support the validity of predicting the rate of future offending, the rates of offending for low-rate and high-rate offenders are not significant enough to support a policy of selective incapacitation. Appendixes discuss the details of the mathematical analysis, and tabular data and a list of 16 references are provided.