An accident involving pedestrian may easily result in fatal injury due to their vulnerability. As reducing fatality in road accident becomes the road safety target, effort should be taken to identify fatality contributing factors through analysis of existing accident data. Thus, this study aims to investigate the factors associated with pedestrian fatal injury. Utilising pedestrian accident data from the year 2009-2013, a model was developed using logistic regression to identify the factors related to pedestrian fatalities at cross and T/Y junctions. The final model explained four significant factors including the age of pedestrian, part of body injury, vehicle and junction type. The results show that accident involving pedestrian occurred at T/Y junction, pedestrian hit by heavy vehicle, and experiencing head/neck injury increase the likelihood of being fatal. Traffic ban for heavy vehicles and speed control at the area with a high volume of pedestrian should be considered due to the greater impact of these factors on the pedestrian fatality.