In recent years tourism has become one of the fastest growing sectors of the world economy and is widely recognised for its contribution to regional and national economic development. The main aim of the paper is to propose a methodological framework within which the impact of characteristics of a tourism product on foreign travel can be captured and analysed. This is achieved by combining the Lancasterian product characteristics approach with the Koppelman's consumer transportation model. Within the proposed model the characteristics of the tourism product/destination including quality of service, advertising and political instability are combined to generate a perception/feeling regarding the tourism destination in the mind of the tourists. In order to measure these perceptions/feelings and, hence, develop a preference ordering, the model is estimated by utilising a conditional Logit analysis. The analysis is based on primary data collected from tourists visiting Cyprus. The conditional Logit model generates the probability of revisit given the characteristics of tourists and the Cyprus tourism product. The paper extends further the Lancaster product characteristics framework as applied to tourism by Rugg (Rev. Econom. Stat. 55(1) (1973) 64), Morley (Ann. Tourism Res. 19 (1992) 250) and Papatheodorou (Ann. Tourism Res. 28 (2001) 164) by including the neglected but from any viewpoint important political instability characteristic that is attached to the tourist product of Cyprus. The inclusion of product characteristics/attributes in tourism analysis appear to strongly contribute towards the better understanding of travel choice behaviour.