This paper assesses the temporal stability of trip generation, distribution, and assignment process at a state or provincial level. The assignment used data from a test study area that included the southwest part of the Province of New Brunswick with a total area of 30 000 km2. For the test year of 1979, trip generation data were developed using three models. The models were based on stepwise regression, principal components, and ridge regression. Following the trip generation phase, a gravity model was used to distribute the output of each model. Given the limited number of alternative routes, an all-or-nothing assignment was used. The stability of the models was assessed using data for 1971. Of the three models evaluated only the one based on ridge regression proved to be reasonably stable over the term period. A stepwise regression model without intercept was also developed and reasonable results were obtained for the stability of this model.Key words: planning, model, stability, trip generation, distribution, assignment, provincial.