This paper develops and estimates a dynamic equilibrium model of the market for commercial aircraft. Airline choices are modeled as the solution to a discrete time dynamic programming problem, where in each period, each airline chooses one or more of various models of new and used aircraft. In equilibrium aircraft prices are such that no airline would benefit from buying, selling, trading or scrapping aircraft. The parameters of the model are estimated by maximum simulated likelihood using a new dataset that contains all aircraft transactions made in the twenty-year period 1978-1997. The transaction data is merged with a dataset containing aircraft prices. The estimated model is used to show that a 10 percent investment tax credit on the purchase of new aircraft has only a small effect on airline behavior and that the demand for new durable goods is more elastic than previous studies have shown. In addition, forcing the modernization of older aircraft causes U.S. airlines to reduce the number of older aircraft they operate by approximately 4 percent, and it is shown that the new aircraft are the poorest substitute for older aircraft.