BACKGROUND: This study investigated the relationship between apparent temperature (AT) and the incidence of acute excessive drinking in Shenzhen, China, and estimated attributable risk fraction and absolute excess number for different apparent temperature ranges and subgroups. METHODS: We conducted a time-series analysis of data on the daily incidence of acute excessive drinking from 2013 to 2017 using a Poisson generalized linear regression model combined with a distributed lag non-linear model. Subgroup analysis by gender and age was also conducted. RESULTS: This study included 85,833 acute excessive drinkers. Both high and low AT showed significantly non-linear and delayed risk effects on excessive drinking; high AT showed acute and strong effects and low AT showed delayed and mild effects. The total attributable risk fraction (AF) contributed by non-optimum ATs was 10.15% (95% empirical CI [eCI]: 5.32-14.38); the AF was the highest in mild heat (75th-95th percentiles) (4.21%, 95% eCI: 1.74-6.30). The absolute excess number (EN) of excessive drinkers were the highest in extreme cold (²1st percentile) and extreme heat (³99th percentile) (205, 95% eCI: 39-335 and 356, 95% eCI: 175-509, respectively). CONCLUSIONS: Both high and low AT had the positive association with the risk of acute excessive drinking. Extreme ATs led to the greatest excess number of acute excessive drinkers.