As factors affecting airport passengers' and other travelers' transportation choices constantly change, the probability of choosing a particular mode also changes. For example, an increase in highway congestion would tend to shift auto travelers to public transport, or an increase in transit fare would make auto traveling more attractive. Although current modal share of airport ground access can be easily determined from passenger's revealed preference surveys, any forecast on mode choice is difficult because characteristics of individual travelers vary. This paper attempts to forecast airport ground access and egress trips at John F. Kennedy International Airport (JFK) with the incremental logit or pivot point model. The model is developed by using modal share data from a revealed preference survey conducted by the Port Authority of New York and New Jersey, which operates JFK. The resulting model predicts how changes in highway travel time due to traffic congestion, increases in transit fare to the airport, and changes in waiting and transfer time in the AirTrain system may affect the mode choice of air passengers and employees traveling to and from the airport.