Abstract
This paper reports on a crash prediction model that was used to estimate the change in truck‐related crashes on parallel arterials caused by the diversion of truck trips from prospective truck‐only toll (TOT) lanes in the Atlanta freeway network. A negative binomial regression was used to model truck crash prediction on arterial roads. Three variables, vehicle‐kilometres travelled, truck percent‐kilometres per lane, and the volume‐to‐capacity ratio, are shown to have significant influence on truck‐related crashes. Two toll policies of mandatory and voluntary use of TOT lanes were examined to determine their effects on truck‐related crashes on parallel arterials. It was estimated that an average of 8% and 6% reduction in truck crash frequency, respectively, would occur.