Accident prediction modeling is a powerful tool for determining the frequency of accidents under certain circumstances. Nationwide, direct damages from highway hazardous material spills for the year 2000 were tallied at over $31billion. This thesis determines the probability of an accident involving hazardous materials on the roads of New Jersey. The methodology is based on a British predictive equation used by their Highways Agency to determine the probability of a hazardous spill over a section of a roadway. The parameters used by the British's Highways Agency, which is obtained from their accident data, were modified to reflect conditions that best fit the State of New Jersey.

Using the probability calculated from this method, the recurrence interval is determined. The recurrence interval represents the number of years it would take before a hazardous material accident would occur. Based on the recurrence interval, segments with higher chances of accidents involving hazardous materials are identified. Thus, by identifying the danger-prone segments, best suited engineering solutions that could be applied to those segments to either arrest spills due to such accidents, or divert them to appropriate places can be made available. This approach would not only benefit the environment efficiently, but would also create fewer disturbances to the public during any hazardous material truck accident.