Abstract
This paper expands an earlier analysis of the effect of firm characteristics and safety practices on truck accident rates. The sample size has been increased from 13, 000 to 75, 500. Negative binomial regressions are used in preference to the Poisson technique used previously. The current analysis confirms previous results, but providesimportant new insights into the safety implications of being a private carrier and hauling hazardous materials and the effect of havingbeen in business for many years. (A).