Resumen
The objectives of the study are to analyse fatal crash patterns, and to identify the factors contributing to motorcycle fatal crashes. An attempt is made to establish relationship between motorcycle versus non-motorcycle fatal crashes and probable contributory factors for three National Highways in India using binomial logistic regression. The distribution of victims? mode by striking vehicles shows that per cent share of striking vehicles as truck against the victims? vehicles, i.e., motorcycle are 44%, 52% and 37% on two-lane, four-lane and six-lane NHs respectively. The results of the model (motorcycle versus non-motorcycle fatal crashes) show that for variable ?collision type?, likelihood to occur rear-end, sideswipe and head-on collision are forty-two times, thirty-five times and twenty-five times more than hit pedestrian respectively; for variable ?number of vehicle?, likelihood is thrice as ?single vehicle? than ?two or more vehicles?; and, for variable ?number of lane?, probability is more on two-lane NH than four-lane NH. To reduce the likelihood of motorcycle fatal crashes especially head-on, rear-end and sideswipe collisions on NHs, it is recommended that riders including pillion riders? use properly fastened standard helmets and compulsorily use of reflective tapes on motorcycles to make them more conspicuous in night.