Safety and accident issues are considered important problems in the world. Road accident issues would have more conspicuous countenance in Malaysia. Almost every day there will be news about road accident, whether on television, newspapers and internet. Therefore, the aim of this study is to understand the general trend of road accidents in Malaysia, to estimate the best parameters of the Box-Jenkins Model and then to find the best-fitted model for road accidents in Malaysia. A monthly accident data from 2007 to 2016 was obtained form the Malaysia Institude of Road Safety (MIROS). This study applies Autoregressive Integrated Moving Average (ARIMA) time series model to study the trend of road accidents in Malaysia. Based on the result obtained, the trend of the number of road accidents in Malaysia are increasing. The finding also showed that by comparing the AIC and BIC value, ARIMA(1,1,1) was identified as the best model. This study contributes to the implementation of road safety in order to reduce the increasing trend of road accidents in Malaysia.