There are two main objectives in the current study: (1) to provide a better understanding of the drunk driving problem in Macau; and (2) to study whether there is a seasonality of drunk driving in Macau. Ninety-six months drunk driving data from 2008 to 2015 were obtained from the monthly traffic reports in Public Security Police Force website. The failed numbers of car drivers involved drunk driving in police roadblocks were analyzed using zero-truncated negative binomial regressions. Besides, sinusoidal equations derived from periodogram ordinates calculations were inserted into the models in order to perform seasonality analysis. The prevalence of drunk driving was about 2% during the entire study period while a decreasing trend was observed in the annual prevalence rate, from 5.41% in 2008 to 1.04% in 2015. Results indicated that models with sinusoidal equations showed a better model fitting against the periodical fluctuations. There are three key findings in this paper: (1) drunk driving appeared to oscillate in a 24-month period cycle in Macau; (2) the failed numbers increased in the first half of a year and start to decreased after July; (3) the failed numbers decreased significantly after February 2011. Practical implications and limitations of the current study were discussed.