Abstract
Multiple-regression analysis of trip generation based on data aggregated to the zonal-average level is discussed with respect to its descriptive and predictive accuracy. The validity of the major assumptions underlying this methodology is examined.