BOTH EVALUATIONS AND SPECIALIZED STUDIES ARE INCLUDED. IN GENERAL, MOST HAVE BEEN PUBLISHED SINCE THE EXAMINATION OF RESEARCH WHICH THE NATIONAL CENTER FOR STATE COURTS ISSUED IN 1975. EVALUATION STUDIES ARE GROUPED ACCORDING TO TYPE OF DESIGN: EXPERIMENTAL, QUASI-EXPERIMENTAL, AND BEFORE-AFTER. RETROSPECTIVE STUDIES USING NO COMPARISON ARE NOT CITED BECAUSE THEY ARE NOT CONSIDERED EVALUATIONS BUT RATHER RESULTS FROM A PROGRAM'S MANAGEMENT INFORMATION SYSTEM. ONLY FOUR EXAMPLES OF TRUE EXPERIMENTAL DESIGN ARE FOUND, BUT EXAMINATION OF THESE DEMONSTRATES THE FEASIBILITY OF THE APPROACH. TWO STUDIES USE A QUASI-EXPERIMENTAL DESIGN. THIS IS CALLED A VALID APPROACH FOR CITY AND REGIONAL EVALUATIONS. THE VALIDITY OF THE LARGE NUMBER OF BEFORE-AFTER STUDIES EXAMINED DEPENDS UPON THE CARE USED IN DRAWING UP THE BASIC RESEARCH DESIGN. SPECIALIZED RESEARCH IS GROUPED ACCORDING TO TOPIC: THE NATIONAL PRETRIAL RELEASE SYSTEM; POINT SCALE, NOTIFICATION, AND SUPERVISION STUDIES; PREDICTOR STUDIES; LOCAL SYSTEM-WIDE STUDIES; AND BONDSMEN. MOST OF THIS WORK FINDS THAT SUPERVISED RELEASE DEFENDANTS HAVE LOWER VIOLATION RATES AND HIGHER COURT APPEARANCE RATES THAN DEFENDANTS RELEASED ON MONEY BOND. HOWEVER, SUPERVISED RELEASE PROGRAMS TAKE LOWER-RISK DEFENDANTS. IT IS RECOMMENDED THAT FUTURE RESEARCH BE DONE INTO CHARACTERISTICS OF PERSONS CHOSEN FOR PRETRIAL RELEASE, WHICH ASPECTS OF THE PROGRAM CAN BE TRANSFERRED TO THE ENTIRE SYSTEM TO IMPROVE PERFORMANCE (FOR INSTANCE, NOTIFICATION OF COURT APPEARANCE OR MORE FREQUENT PERSONAL CONTACT), AND DEGREE OF SUPERVISION WHICH PRODUCES THE BEST RESULTS AT LEAST COST. THE BAIL BOND SYSTEM ALSO NEEDS MORE STUDY. THE NORMATIVE LITERATURE DECRIES THE ABUSES OF THE BONDING INDUSTRY, BUT CASE STUDIES DOCUMENTING THE SYSTEM NEED TO BE COLLECTED. A GLOSSARY OF TERMS IS INCLUDED. CITATIONS FOR ALL 43 STUDIES MENTIONED ARE INCLUDED IN DETAILED FOOTNOTES.