This publication reports the results of the first study in a research series aimed at investigating the development of a model of the recidivism rate that would be valid for several years, using data available on the Offender Records System (ORS), and at identifying high and low recidivism groups in the total inmate population using data available on the ORS. Data used was for all inmates discharged in 1990 or 1991, after serving a full- time adult custodial sentence in New South Wales. Methodology is explained in detail. Other recidivism studies are covered briefly. Results of this study indicate that, in general, the following variables available on the ORS are related to recidivism: prior adult imprisonment, age, time between the focal episode (as defined in the study) and prior imprisonment episode, if any, security classification at discharge, and the most serious offense of focal episode. The relationship varied for people discharged at different times. Some groups with high or low overall recidivism can be identified for special targeting using variables available on the ORS. Not all inmates in each group, however, had a corresponding high or low probability of recidivism. Tables, references