THREE OBJECTIVES OF THIS STUDY WERE CHOSEN: (1) TO FIND DEMOGRAPHIC VARIABLES WHICH SIGNIFICANTLY AFFECT THE PROBABILITY THAT AN ACCUSED WILL NOT APPEAR IN COURT, (2) TO DETERMINE THE RELATION BETWEEN BOND PRICE AND THE SIGNIFICANT VARIABLES FOUND IN OBJECTIVE ONE, AND (3) TO COMPARE THE SIGNIFICANT VARIABLES FOUND IN OBJECTIVE ONE TO THE CRITERIA USED FOR DENYING RELEASE IN THE DALLAS COUNTY PRE-TRIAL RELEASE PROGRAM. IT WAS HYPOTHESIZED THAT AN ACCUSED PERSON WOULD USE A COST-BENEFIT APPROACH WHEN DECIDING WHETHER TO APPEAR FOR TRIAL OR FORFEIT THE BAIL BOND, AS WELL AS AN ASSESSMENT OF RISK OF DETECTION. FIVE VARIABLES PROVED TO BE GOOD PREDICTORS OF BAIL BOND FORFEITURE: (1) RESIDENCE OUT OF COUNTY, (2) CHARGED WITH THEFT, (3) CHARGED WITH WEAPONS OFFENSE, (4) CHARGED WITH NARCOTICS OFFENSE, AND (5) NUMBER OF PREVIOUS ARRESTS. HOWEVER, THESE VARIABLES HAD VIRTUALLY NO EFFECT ON THE SETTING OF THE BOND PRICE, NOR DID THE PRICE CORRELATE WITH TRIAL APPEARANCE. IN THE EVALUATION OF THE DALLAS COUNTY PRE-TRIAL RELEASE PROGRAM IT WAS FOUND THAT THE SIGNIFICANT PREDICTOR VARIABLES WERE ALL USED BY THE PROGRAM TO EXCLUDE ACCUSED PERSONS FROM RELEASE (FOR THE NOMINAL FEE OF $15). THE PROGRAM'S CRITERIA FOR DENYING RELEASE WAS FOUND TO BE MORE COMPREHENSIVE AND RESTRICTIVE THAN THE LIST OF FIVE PREDICTOR VARIABLES UNCOVERED IN THE EARLY PART OF THE ANALYSIS. IT IS NOTED THAT ONLY LOW-RISK PERSONS WERE RELEASED THROUGH THE PRETRIAL RELEASE PROGRAMS, WHICH WEEDED THESE PERSONS OUT BEFORE REACHING THE BAIL-SETTING PHASE OF THE PROCESS. TABLES AND NOTES ARE PROVIDED. (MLC)