THE MODEL EMPLOYS THE TECHNIQUE OF PATH ANALYSIS AND PARTIAL REGRESSION COEFFICIENTS TO ASCERTAIN EFFECTS OF CERTAIN INDEPENDENT VARIABLES ON THE PROBABILITY OF DRUNKEN DRIVING, THE DEPENDENT VARIABLE. A QUESTIONNAIRE WAS MAILED IN THE SPRING OF 1975 TO A RANDOM SAMPLING CONSISTING OF 3,000 PERSONS BETWEEN 15 AND 60 YEARS OF AGE, AND 1,541 RESPONDENTS OR 73 PERCENT OF THE SAMPLE PROVIDED DATA FOR ANALYSIS. INDEPENDENT VARIABLES PRESUMED TO AFFECT THE PROBABILITY OF DRUNKEN DRIVING WERE AGE, SEX, MARITAL STATUS, ALCOHOL CONSUMPTION, MILEAGE, PERCEPTION OF ACCIDENT RISK, KNOWLEDGE OF THE LAW, MORAL ATTACHMENT TO THE LAW, AND HISTORY OF DRUNKEN DRIVING. THE RELATION BETWEEN SEX AND DRUNKEN DRIVING IS EXPLAINED BY MEN'S HIGHER ALCOHOL CONSUMPTION, HIGHER MILEAGE, AND LOWER ESTIMATION OF ACCIDENT RISK. THE IMPACT OF AGE AND MARITAL STATUS IS TRANSMITTED THROUGH THE SAME MECHANISM AS THAT LINKING THE SEX EFFECT TO DRUNKEN DRIVING. IN GENERAL, THE MODEL INDICATES THAT VARIATIONS IN DEMOGRAPHIC CHARACTERISTICS OF THOSE WHO DRINK AND DRIVE CAN BE EXPLAINED BY THE FACT THAT MEN, YOUNG PEOPLE, AND UNMARRIED PERSONS SCORE DIFFERENTLY FROM THEIR COUNTERPARTS ON VARIABLES DIRECTLY OR INDIRECTLY CONTRIBUTING TO DRUNKEN DRIVING. ALCOHOL CONSUMPTION AND MILEAGE STAND OUT AS THE MOST CRUCIAL FACTORS CONTRIBUTING TO DRUNKEN DRIVING. DRIVING UNDER THE INFLUENCE OF ALCOHOL IN SWEDEN IS KNOWN GENERALLY TO BE ILLEGAL, AND THE FINDINGS OF ANALYSIS DEMONSTRATE THAT MORE KNOWLEDGE CONCERNING ALCOHOL-RELATED OFFENSES AND RISKS DOES NOT AFFECT THE TENDENCY TO DRINK AND DRIVE. THE TENTATIVE NATURE OF THE MODEL AND ITS LIMITATIONS ARE DISCUSSED. SUPPORTING DATA, A PATH DIAGRAM, AND NOTES ARE INCLUDED. (DEP)