MARTINSON CONCLUDED IN HIS RESEARCH THAT, WITH FEW AND ISOLATED EXCEPTIONS, REHABILITATIVE EFFORTS HAVE HAD NO APPRECIABLE EFFECTS ON RECIDIVISM. TWO POINTS ARE CENTRAL TO UNDERSTANDING MARTINSON'S DECISION TO ABANDON RECIDIVISM RESEARCH IN FAVOR OF CRIME RATE RESEARCH. FIRST, HE BELIEVES THAT CRIME RATE STUDIES PROVIDE MORE VALUABLE INFORMATION AND THAT THE RECIDIVISM MEASURE IS UNIMPORTANT. SECOND AND IMPLICIT IN THE FIRST ARGUMENT, RECIDIVISM MEASURES AND OVERALL CRIME STATISTICS MUST PROVIDE DIFFERENT OR UNRELATED INFORMATION ABOUT CRIMINAL BEHAVIOR. MARTINSON'S INDIFFERENCE TO RECIDIVISM IS BASED ON HIS PERSONAL SENSITIVITY TO PUBLIC ATTITUDES. AN ESSENTIAL COMPONENT IN THE DEVELOPMENT OF A THEORY OF CRIMINALITY INCLUDES THE SCIENTIFIC INVESTIGATION OF REPETITIVE CRIMINALITY. THE TOTAL CRIME PICTURE IS A COMPOSITE OF REPORTED AND UNREPORTED CRIMINAL ACTS. SINCE THE LIKELIHOOD OF INCARCERATION IS PROPORTIONAL TO THE FREQUENCY OF COMMITTING CRIMINAL ACTS, IT FOLLOWS THAT MANY OF THE SMALL PERCENTAGE OF PERSONS WHO ACCOUNT FOR A LARGE PORTION OF TOTAL CRIME IN SOME COMMUNITIES EVENTUALLY FIND THEIR WAY INTO THE PRISON SYSTEM. RECIDIVISM STATISTICS SUPPORT THIS LINE OF REASONING. THERE IS NO REASON TO SUSPECT THAT PROCESSES MEDIATING INCARCERATION-REINCARCERATION ACTIVITIES WILL BE UNRELEATED TO PROCESSES MEDIATING ANY OTHER KIND OF CRIMINAL ACTIVITY. PROCESSES MEDIATING CHANGE IN CRIMINAL DISPOSITION ARE DISCUSSED, AND DEFICIENCIES IN MARTINSON'S THESIS ARE NOTED WITH REGARD TO SCIENTIFIC INQUIRY, DATA GENERATION, AND THEORY CONSTRUCTION. REFERENCES ARE CITED. (DEP)