THE U.S. DEPARTMENT OF TRANSPORTATION POSED TWO RESEARCH QUESTIONS: (1) WHETHER MANDATORY COURT APPEARANCE IS A MORE EFFECTIVE MEANS OF HANDLING VIOLATIONS OF ORDINARY TRAFFIC LAWS THAN APPEARANCE AT A VIOLATIONS BUREAU, WITH THE ROUTINE GUILTY PLEA AND PAYMENT OF A SCHEDULED FINE; AND (2) WHETHER PROBATION (OR PROBATION COMBINED WITH EDUCATIONAL OR CLINICAL PROGRAMS) IS BETTER THAN MONETARY FINES AS A MEANS OF HANDLING DRIVERS CONVICTED FOR THE FIRST TIME OF DRIVING WHILE INTOXICATED. A PRELIMINARY STUDY SUGGESTED THAT RANDOM ASSIGNMENT OF DRIVERS TO EXPERIMENTAL CONDITIONS WAS DESIRABLE. ETHICAL QUESTIONS INVOLVING THE ASSIGNMENT OF INDIVIDUALS TO LEGAL PUNISHMENT ON OTHER THAN A LEGAL BASIS AROSE. TWO SOMEWHAT OPPOSING PRINCIPLES IN LAW REGARDING THE PROPER BASIS FOR SENTENCING -- THAT THE PUNISHMENT FIT THE CRIME AND THAT THE PUNISHMENT FIT THE CRIMINAL -WERE CONSIDERED. A STUDY DESIGN, DEVELOPED IN LIGHT OF THESE PRINCIPLES, WAS JUSTIFIED ETHICALLY AS FOLLOWS: ALL PRESCRIBED TREATMENTS WERE COMMONLY USED IN THE COURT SYSTEM UNDER STUDY FOR THE TYPES OF CASES TO WHICH THEY WERE APPLIED IN THE EXPERIMENT; IN THE DRINKING AND DRIVING STUDY, THE EXPERIMENT GOVERNED ONLY THE QUALITY OR TYPE OF SANCTION, NOT THE AMOUNT; IN THE MOVING VIOLATION STUDY, THE DIFFERENCE BETWEEN THE GREATEST AND LEAST INCONVENIENCE EXPERIENCED BY SUBJECTS WAS SLIGHT; AND THE VALUE OF THE POSSIBLE RESULTS OF THE STUDY SEEMED TO OUTWEIGH ANY MARGINAL COSTS TO THE CITIZENS INVOLVED. THE EXPERIMENTERS' PRACTICAL PROBLEMS INCLUDED UNEXPECTED, SIGNIFICANT DIFFERENCES BETWEEN EXPERIMENTAL AND CONTROL GROUPS IN THE MOVING VIOLATION STUDY, AND DEVIATIONS FROM EXPERIMENTAL DESIGN BY POLICE AND COURT PERSONNEL, PARTICULARLY JUDGES. RESEARCHERS SHOULD NOT EXPECT THE PROMISES OF COOPERATION FROM LEADERSHIP TO GUARANTEE COOPERATION BY 'FRONT LINE' PERSONNEL. WHERE POSSIBLE, DECISIONMAKING ABOUT ASSIGNMENT OF SUBJECTS TO TREATMENT GROUPS SHOULD REST WITH EXPERIMENTERS AND NOT BE DELEGATED TO BUREAUCRATIC PERSONNEL. ASSIGNMENT DECISIONS THAT MUST BE MADE BY BUREAUCRATIC PERSONNEL SHOULD BE MONITORED CONSTANTLY. WHEN PRESSURE BY CLIENTS (I.E., THE ACCUSED) TO DEVIATE FROM RESEARCH PRESCRIPTIONS IS EXPECTED TO BE INTENSE, THAT PRESSURE SHOULD BE ALLOWED TO HAVE ITS EFFECT BEFORE THE DECISION AFFECTING EXPERIMENTAL VALIDITY IS MADE. A LIST OF REFERENCES IS INCLUDED. (LKM)