This study is to investigate the relation of freeway drunk driving accidents to law enforcement for driving while intoxicated, social and economic factors by using analysis of correlation and regression analysis. As shown in the results of the study, law enforcement for driving while intoxicated, the overall law enforcement per week/month is significantly negatively correlated with the overall drunk driving accidents for the same week/month and the next week/month. In terms of a whole year, law enforcement for driving while intoxicated per year is significantly negatively correlated with the incidence rate of Class A1 drunk driving accidents for the same year and the next year. Therefore, the police's operation of law enforcement has considerable control and preventive effect upon the incidence rate of drunk driving accidents but also a great influence upon people's future attitudes towards law abidance. As for social and economic factors, national annual incomes and vehicle ownership rates are significantly negatively correlated with the incidence of drunk driving accidents, it is suggested that law enforcement for driving while intoxicated should continue to implement, and other preventive and control measures such as education, propaganda, and legislation should be taken to effectively curb drunk driving accidents.