DATA SOURCES FOR THE EVALUATION WERE INTERVIEWS WITH 732 CLIENTS AND RECORDS OF 1,633 CLIENTS. PRIMARY GOALS OF THE CONDITIONAL RELEASE PROGRAM, WHICH BEGAN IN SEPTEMBER 1973, ARE TO REDUCE THE DETENTION CENTER POPULATION, INCREASE THE PROBABILITY THAT THE RELEASEES WILL RETURN FOR TRIAL, AND DECREASE RELEASEES' REARREST RATES. SOCIAL BENEFITS ARE ALSO ANTICIPATED. THE 3-YEAR EVALUATION OF THIS PROJECT BEGINNING OCTOBER 1973 REVOLVED AROUND THE PROJECT'S ABILITY OR INABILITY TO MEET ITS GOALS. IT WAS FOUND THAT THE CONDITIONAL RELEASE PROGRAM WAS NEVER FULLY IMPLEMENTED IN ACCORDANCE WITH ITS PROPOSAL. THE EVALUATION STATES THAT THE PROJECT HAS NOT MET ANY OF ITS OPERATIONAL GOALS. INSTEAD OF AN ACTIVE CASELOAD OF 1,500 CASES, IT HAS CONSISTENTLY SERVED LESS THAN 200 CASES AT A TIME. INSTEAD OF RELEASING DEFENDANTS WITHIN 5 DAYS OF ARREST, IT AVERAGES 10 DAYS. INSTEAD OF MAINTAINING DEFENDANTS IN THE PROGRAM FOR 120 DAYS, IT MAINTAINS THEM FOR 75 DAYS. IT WAS FOUND THAT THE PROGRAM DID EFFECT A REDUCTION IN THE DETENTION CENTER POPULATION IN PHILADELPHIA. THE PROGRAM ALSO DOES NOT APPEAR TO INCREASE THE RISK TO THE COMMUNITY MORE THAN OTHER PRETRIAL RELEASE PROGRAMS. THE FAILURE TO APPEAR AND REARREST RATES FOR CLIENTS IN THE CONDITIONAL RELEASE PROGRAM ARE LOWER THAN FOR COMPARABLE PROGRAMS. COMPARED TO OTHER ALTERNATIVES, THE CONDITIONAL RELEASE PROGRAM IS LESS COSTLY. COMMUNITY AGENCIES, JUDGES, AND DEFENDANTS ARE ALL FAVORABLY DISPOSED TOWARDS THE PROGRAM, ALTHOUGH THERE IS NO EVIDENCE TO INDICATE THAT CONDITIONAL RELEASE IMPROVED DEFENDANTS' SELF-CONCEPTION. FIGURES, TABLES, AND DATA COLLECTION INSTRUMENTS ARE APPENDED. (AUTHOR ABSTRACT MODIFIED--MLC)