RESULTS SUPPORT THE GLASER FINDING THAT THE LONGITUDINAL RECIDIVISM RATE IS CLOSER TO ONE-THIRD (38 PERCENT IN ILLINOIS AND 39 PERCENT IN WISCONSIN) THAN TO THE FREQUENTLY CITED TWO-THIRDS RATE WHEN RECIDIVISM IS MEASURED OVER THE CONVENTIONAL 3-YEAR FOLLOWUP PERIO. ALSO SUPPORTED IS THE GLASER FINDING THAT CROSS-SECTIONAL STUDY YIELDS A TWO-THIRDS RECIDIVISM RATE (72 PERCENT IN ILLINOIS). INCOMPLETE ILLINOIS DATA SHOWED THE RECIDIVISM RATE INCREASING FROM 38 PERCENT AFTER 3 YEARS TO ALMOST 50 PERCENT AFTER 6 YEARS, SUPPORTING THE KITCHENER, SCHMIDT, AND GLASER FINDING THAT THE RECIDIVISM RATE DOES NOT BECOME STABLE AFTER 3 YEARS, BUT CONTINUES TO RISE FOR ABOUT 8 YEARS. ILLINOIS RESULTS EXTEND THE FINDING THAT CRIMINAL HISTORY PREDICTS RECIDIVISM. THE TRADITIONAL BELIEF THAT ADULT OFFENDERS WERE JUVENILE OFFENDERS IS NOT SUPPORTED (ONLY 27 PERCENT IN ILLINOIS HAD A JUVENILE OFFENSE HISTORY). RESULTS SUPPORT THE FOLLOWING CONCLUSIONS: (1) THE LONGITUDINAL/CROSS-SECTIONAL DISTINCTION IN MEASURING RECIDIVISM RATES IS RELEVANT TO CORRECTIONS FOR POLICY AND PLANNING PURPOSES; (2) THE APPROPRIATE LENGTH OF THE FOLLOWUP PERIOD FOR DETERMINING RECIDIVISM SHOULD BE REEXAMINED; (3) RECIDIVISM RATES ARE SENSITIVE TO MEASURES OF PRIOR CRIMINALITY WHILE BEING INSENSITIVE TO OFFENDER PROGRAM DIFFERENCES; AND (4) RECIDIVISM RATES FOR DIFFERENT PRISON SYSTEMS APPEAR TO BE A REFLECTION OF THE TYPES OF OFFENDERS SELECTED FOR IMPRISONMENT RATHER THAN THE KINDS OF PROGRAMS OFFERED. TABULAR DATA AND FOOTNOTES ARE PROVIDED. (AUTHOR ABSTRACT MODIFIED--RCB)