Background Predicting the death trend caused by traffic accidents and its analysis is a useful tool for planning and policy-making by the authorities in the field of road traffic, conducting interventions appropriate with death trend, and taking the actions required for controlling future conditions.

Methods In a cross-sectional research, all the information related to the traffic accidents leading to death available in the database of Iran Legal Medicine Organisation from 2004 to the end of 2013 was used to determine the change points (multi-variable time series analysis). Role of regulations and rules in the variations of death caused by traffic load was studied using Box Jenkins multi-variable time series models and critical path analysis (CPA) over time. Using autoregressive integrated moving average (ARIMA) model, rate of death caused by traffic accidents was predicted for 2014 and 2015 and, finally, the actual rate of death caused by traffic accidents was compared with the predicted value to determine the efficiency of the model.

Results Results showed that rate of death in 2014 was almost equal to the actual rate of death recorded for this year, while rate of death in 2015 decreased compared with the previous year (2014) during all months. It was also predicted that, in January and February 2015, rate of death would reach its minimum value in that year, i.e. 4.1%.

Conclusions According to the prediction and analysis of the trend of death caused by traffic accidents, it was observed that applying and continuing the intervention conducted in the previous years for road safety improvement, motor vehicle safety improvement, particularly training and culture-fostering interventions, as well as approval and execution of deterrent regulations for changing the organisational behaviours can significantly decrease the loss caused by traffic accidents.