Passenger preference analysis is an important area of transport modeling. Passengers value public transport journeys in several ways: their opinion depends on many factors. The relative importance of these features is often characteristic to certain passenger groups. Knowing the distribution of the passengers and their preferences makes it possible to improve transport models and simulation techniques. The paper provides an overview of existing revealed and stated preference analysis methods and the results of some former surveys carried out in Gyor, Hungary. Then it proposes a new method for detecting passenger preferences by integrating data collection into an online journey planning service.