["Addiction, Volume 121, Issue 5, Page 1166-1178, May 2026. ", "\nAbstract\n\nBackground and aims\nDrunk‐driving accounts for a significant portion of traffic crashes in Taiwan. From 2008 to 2013, both the Taiwan Legislative Yuan and the federal administrative agencies collaborated to increase drunk‐driving penalties. This study aimed to determine whether these sanction changes are associated with changes in the incidence of drunk‐driving and subsequent hospital admissions.\n\n\nDesign\nExploratory retrospective cohort study.\n\n\nSetting\nTaiwan.\n\n\nParticipants\nAll traffic crashes involving drunk‐driving violations and every involved individual from 2003 to 2019 in Taiwan.\n\n\nMeasurements\nInterrupted time series analysis (ITSA) method was used to examine the effects of the three policy interventions in January 2008, December 2011 and June 2013. The numbers of drunk‐driving crashes, drunk‐driving casualties, drunk‐driving‐related admissions, drunk‐driving‐related major traumas, drunk‐driving‐related deaths and drunk‐driving‐related medical expenses in different time segments were compared to evaluate the effect of each policy intervention. All three policy initiatives increased drunk‐driving penalties, and there were no intervening policy changes that reduced said penalties.\n\n\nFindings\nFrom January 2003 to January 2008, the numbers of all drunk‐driving‐related outcomes gradually increased. After the first penalty adjustment, the reported number of drunk‐driving crashes had a sudden decrease (1632 in January 2008 to 1374 crashes in February 2008, β2 = −0.146, P < 0.001), but the overall trend continued upward (β3 = −0.001, P = 0.111). However, after the second policy intervention, drunk‐driving crashes began trending downward rapidly (β5 = −0.021, P < 0.001). Finally, after the third adjustment, the trend of drunk‐driving crashes continued to decrease, but the slope became significantly milder (β7 = 0.009, P < 0.001). Most of the target outcomes followed a similar trend as the drunk‐driving crashes, except for the number of deaths caused by drunk‐driving, which demonstrated a statistically significant decline after the first adjustment (β3 = −0.006, P = 0.007).\n\n\nConclusions\nIncreasing the fine and maximum incarceration periods for drunk‐driving in Taiwan appears to be associated with a reduction of the number of drunk‐driving crashes and other related outcomes; however, the decrease in drunk‐driving crashes diminished over time, especially after the third sanction change in 2013.\n\n"]