DATA WERE OBTAINED DURING THE FIRST QUARTER OF 1977 AT AN ALCOHOLISM TREATMENT CENTER IN THE STATE OF WASHINGTON. THE 101 STUDY SUBJECTS WERE CHOSEN RANDOMLY FROM EACH OF TWO DIFFERENT TREATMENT CONDITIONS: (1) CLIENTS WHOSE SENTENCE FOR DWI HAD BEEN SUSPENDED WITH THE STIPULATION THAT THEY UNDERGO ANTABUSE (ALCOHOL AVERSIVE) TREATMENT AND (2) CLIENTS WHO WERE BEING TREATED WITH ANTABUSE BUT WHO HAD NOT BEEN REFERRED BY COURTS. STATISTICAL ANALYSIS TECHNIQUES WERE EMPLOYED TO DETERMINE THE RELATIVE EFFECTIVENESS OF EACH TREATMENT CONDITION ON DWI RECIDIVISM FOR A 1-YEAR PERIOD AFTER EACH CLIENT HAD COMPLETED THE ANTABUSE TREATMENT PROGRAM. THE LOWER RATE OF DWI CLIENTS IN THE FIRST TREATMENT CONDITION WAS ASSOCIATED WITH THE COURT-THREATENED PENALTY OF A JAIL SENTENCE FOR NOT PARTICIPATING IN THE ANTABUSE TREATMENT PROGRAM. FURTHER RESEARCH IS RECOMMENDED TO PROVIDE A RATIONAL JUSTIFICATION FOR POLICY DECISIONS ON THE USE OF LEGAL COERCION IN ALCOHOLISM TREATMENT. COMBINING THE CRIMINAL JUSTICE SYSTEM WITH THE ALCOHOLISM TREATMENT COMMUNITY IS DISCUSSED. TABULAR DATA AND FOOTNOTES ARE INCLUDED.