THE STUDY WAS CONDUCTED IN SEATTLE, WASH., OVER A PERIOD OF 6 MONTHS. PUBLIC DRUNKENNESS ARRESTS WERE A MAJOR PERCENTAGE OF ALL CRIMINAL ARRESTS IN THE 5 YEARS PRECEDING THE STUDY, AND SEATTLE EXEMPLIFIED THE TRADITIONAL LEGAL STRATEGY OF PUBLIC DRUNKENNESS CONTROL. INTOXICATED PERSONS IN A PUBLIC PLACE WERE SUBJECT TO ARREST UNDER A MUNICIPAL ORDINANCE PROHIBITING DRUNKENNESS. DATA WAS COLLECTED DURING 8 MONTHS OF PARTICIPANT OBSERVATION OF ROUTINE PATROLS. THE UNIT OF ANALYSIS WAS THE DRUNKENNESS CONTROL ENCOUNTER (AN ENCOUNTER BETWEEN POLICE AND AN INTOXICATED CITIZEN). ABOUT 740 ENCOUNTERS WERE OBSERVED AND INFORMATION GATHERED ON WHO INITIATED THE ENCOUNTER; AGE, SEX, AND RACE OF SUBJECT, HIS DEGREE OF INTOXICATION, SIGNS OF INJURY OR ILLNESS, DEMEANOR; OTHER CRIMINAL ACTIVITY; WHETHER A CITIZEN COMPLAINANT WAS PRESENT, AND WHETHER THE CASE ENDED IN FORMAL PROCESSING. NATIVE AMERICANS WERE THE SINGLE MOST FREQUENTLY ENCOUNTERED GROUP IN THE SAMPLE WITH REPRESENTATION FAR ABOVE THAT IN THE SEATTLE POPULATION. WHITES, THE SECOND MOST FREQUENTLY ENCOUNTERED GROUP, WERE UNDERREPRESENTED IN TERMS OF THEIR NUMBERS IN THE SEATTLE POPULATION. FINDINGS INDICATED THAT FORMAL PROCESSING WAS HIGHLY RELATED TO NORMATIVE ASPECTS OF THE ENCOUNTER SITUATION. VIOLATION OF FORMAL, LEGAL NORMS INCREASED THE LIKELIHOOD OF ARREST. ARREST WAS ALSO FOUND MORE LIKELY IF A CITIZEN COMPLAINANT WAS PRESENT AT THE ENCOUNTER. FINDINGS ALSO SHOWED THAT 75 PERCENT OF THE ENCOUNTERS OCCURRED THROUGH POLICE INITIATIVE AND THAT ONLY ABOUT 14 PERCENT OF THE ENCOUNTERS ENDED IN FORMAL PROCESSING. THESE SUGGEST THAT THE LAW WILL SUPPORT LESS FORMAL MODES OF CONTROL AND WILL ASSUME THE ROLE OF MORE FORMAL CONTROL ONLY AS A SYSTEM OF LAST RESORT. TABULAR DATA AND A LIST OF REFERENCES ACCOMPANY THE ARTICLE. (MJW)