In this study a cross-classification technique is used to predict trip production travel among 20 traffic analysis zones located within Dohuk city residential area. Two and three level cross-classification matrices has been used to describe dis- aggregated trip rates/Du, total vehicle trips/Du and total private trips/Du. Car ownership is considered as the main factor causing trip production related to other household characteristics like family size income level and workers number.
Out of this study it is concluded that data can be used directly in the prediction analysis of trip rates. Family size and workers number are the most effective independent variables. Number of cells can be reduced if larger sample size are used in the prediction analysis as well.
Key Word : Cross Classification, Dohuk