the effect of petroleum rationing on road traffic fatalities in the south of iran during 2004-2009

background: this study attempted to investigate traffic accident fatalities during 2004-2009 and the effect of fuel rationing on traffic accident fatalities .   materials and methods: this was a cross-sectional study on available data of all traffic accident fatalities in fars province, iran, during 2004-2009. to identify and fit the best model, various instruments, including the autocorrelation function (acf), partial autocorrelation function (pacf), and akaike information criterion (aic), were used. to determine the accuracy of the model, the residual chart was evaluated in terms of independency, normality, and stationary. then, the best model in terms of more accurate estimation was selected and analyzed using itms, minitab, and microsoft excel software.   results: during 2004 to 2009, 12954 people died due to road traffic fatalities in fars province. among them, 10442 (80%) were male, and mean age at death was 36 ± 20. the number of traffic accident fatalities in fars province reduced from 45.5 per 100000 in 2004 to 44.8 per 100000 in 2007 . after petroleum rationing, the traffic accident fatality average further reduced (to 42.2 per 100000 in 2009) . t here was a difference between the observed value and predicted value with two different predication methods, as the number of observed fatalities was less than the estimated numbers. seasonal holt-winters and autoregressive moving average (arma) methods’ predications were a close estimation of the accident fatality rate after petroleum rationing, and the difference between their estimations was not considerable.   conclusions: although the overall rate of fatalities reduced after petroleum rationing, variation in fatality rate did not show any association with petroleum rationing. however, decrease in fatalities had a seasonal pattern. further decrease in road traffic fatalities requires reinforcement of educational programs as well as application of effective laws.