Background The first Road Traffic Safety Law in China had been adopted on Oct 28th, 2003 and implemented on May 1 st, 2004, but there is far less evidence about the effect of this law on road traffic deaths (RTD), especially at subnational level.

Objective We assessed the effect of the law on the RTD before and after the implementation at both national and subnational level.

Methods 2005, as the closest integer year to May 1 st, 2014, is chosen as the start of the intervention to assess the effect. Given the law targeting the entire population, the interrupted time-series method with Newey-West standard error was applied based on the estimated results from Global Burden Disease 2015. Cumby-Huizinga test was used to ensure that all models could correctly account for autocorrelation structure.

Findings There was a significant increasing trend of the RTD in China before 2005, followed by a reversed trend with a significant difference compared with the prior. No significant immediate effect was found after 2005 until 2008. Among 31 provinces, different growth trends occurred in 27 provinces before 2005 and nearly all provinces saw downward trend after 2005 except for Guangxi and Hubei. Significantly decreased immediate effect was found in only 9 provinces after 2005. In other 22 provinces, 15 provinces had a significantly decreased immediate effect with different lagged years from 2006 to 2010.

Conclusion The expected decreasing trend of the RTD has occurred in China and most provinces after the law enforcement, but the effect varied among provinces.

Policy implications The road traffic safety law implemented in China could be a good reference for other developing countries within a similar socio-economic paradigm shift. This study also provides a clue for further research on the reason of disparity of the effect of the law in different provinces.