A FRAMEWORK FOR DWI GENERAL DETERRENCE IS DEFINED; THE ANALYTICAL METHODS EMPLOYED ARE DESCRIBED. CENTRAL TO THE STUDY WAS A SYSTEM MODEL DEVELOPED FOR INTERRELATING FACTORS WHICH INFLUENCE DWI DETERRENCE, AND AN ASSOCIATED COMPUTER-BASED SIMULATION PROGRAM EMPLOYED FOR EXAMINING DWI DETERRENCE ALTERNATIVES. RESULTS OF THE ANALYSIS INDICATED THAT (1) ANY SIGNIFICANT REDUCTION OF DWI TRIPS OR RELATED ACCIDENTS MUST BE EFFECTED THROUGH GENERAL RATHER THAN SPECIFIC DETERRENCE; (2) GENERAL DETERRENCE DEPENDS UPON DRIVERS' PERCEIVED RISK OF DWI TRIPS AND ON RISK AVERSION; (3) RELATIVELY SMALL CHANGES IN PERCEIVED RISK OF DWI ARE LIKELY TO PRODUCE LARGE CHANGES IN NUMBER OF DWI TRIPS OR ACCIDENTS; (4) FEEDBACK FROM APPREHENDED DRIVERS IS UNLIKELY TO RESULT IN ANY SIGNIFICANT REDUCTION IN TRIPS OR ACCIDENTS; (5) INCREASED ENFORCEMENT ACTIONS REDUCE DWI TRIPS AND ACCIDENTS; AND (6) THE GREATEST POTENTIAL FOR REDUCED DWI TRIPS OR ACCIDENTS IS THROUGH DISSEMINATION OF INFORMATION FROM EFFECTIVE AND CONSISTENT DWI ENFORCEMENT AND ADJUDICATION ACTION. REFERENCES ARE INCLUDED. (MJW)