Very few studies have explored residual career length (RCL) and residual number of offenses (RNO), that is, the remaining time and number of offenses in criminal careers. This study uses conviction data from the Cambridge Study in Delinquent Development to investigate RCL and RNO, for a sample of British males and their fathers. The sons were followed up to age 40 and the fathers up to age 70. Distributions of RCL and RNO according to six different criteria are presented (age on offense, conviction number, time since the last conviction, age of onset, offense type, and number of co-offenders). There was a general decline in RCL and RNO with age. Although RCL declined steadily with each successive conviction for both sons and fathers, RNO did not decline with conviction number for fathers. Over and above age on conviction, age of onset predicted RCL and RNO for sons, but less so for fathers. The type of offense and the number of co-offenders did not predict RCL or RNO. Risk scores showed that the predictive power of these variables for RCL and RNO was statistically significant but not very high. This finding highlights the difficulties associated with predictions of criminal career outcomes based on information available in official records, which is the main source of information available to decision-makers in the criminal justice system.