Abstract
As part of economic studies concerning the plans for the construction of 150 miles of motorway in Jutland, Denmark, a model for forecasting differences in traffic accidents for certain construction stages was developed. By way of regression techniques models for the relationships between accidents, highway mileage, cross sections, traffic flow and development along the road were set up. Furthermore, it was checked by way of variance ratio tests (f-tests) whether significant differences in the relationships exist between different parts of the country.