Air transportation traffic is increasing, and this results in newer and larger airport investments in outskirts of the cities. Thus, airport ground access is becoming an important topic to focus on. Indeed, there are many studies in this field covering the mode choices of passengers to and from the airports. In this case study, the mode choices of travelers who are residents of Istanbul for access to Sabiha Gokcen International Airport (SAW) are modeled. To collect data, a revealed-preference survey with passengers is conducted at SAW. A literature review indicated that as well as other factors, trip purpose as business or personal affected the airport ground access mode choice. In accordance with this finding, separate mode choice models for both business and non-business passengers, as well as a pooled model for all travelers are built. Multinomial Logit (MNL) is used to model the mode choices. The modes were automobile, taxi and public transit. The results indicated that trip cost to SAW, traveling group size, time difference between the departure time and flight time, and automobile ownership status of the passenger were the explanatory variables for mode choice model. A market segmentation analysis results further showed that building separate models for business and non-business passengers was an improvement over the pooled mode choice model.