Restitution was defined as court imposed monetary payments to crime victims in compensation for crime-related losses. The offenders were convicted and sentenced in an Oregon circuit court. Less than half were in compliance with court orders at the time of the study. Twelve variables were found to be important in predicting compliance rates. These included offender characteristics, offense and sentence, and restitution orders. Offender typologies were constructed to characterize compliance with restitution orders. Data from four probation courts in Georgia validated these typologies. The effects of imposing and paying restitution upon recidivism while on probation were examined using expanded Oregon circuit court data. The imposition of restitution had no effect on recidivism. Paying restitution was associated with reduced recidivism. The data did not substantiate the rehabilitative, deterrent, retributive, and reparative claims offered by proponents of restitution. There is a need to modify and refine procedures related to the imposition and enforcement of restitution to increase the likelihood that restitution will be paid. Tables and 60 references are included. (Author abstract modified)