Currently, modal split modeling is done mainly by means of disaggregated mode choice models. The almost absolute dominance of multinomial and nested logit models over other mode choice models among applied transportation modelers is attributable to their theoretical soundness, to their simple and understandable analytical structure, and to the calibration procedures that have been developed. Typical urban transport systems, however, are characterized by a variety of modes including private (automobile), public transit (bus, suburban rail, light rail, and subway), and various combinations of these. Analysis reveals that the nested logit model based on the assumption of groupwise similarities among modes is not a suitable modeling tool in such situations. A cross-nested model that is derived from the generalized extreme value class and that can be thought of as a generalization of the nested logit model is proposed. The model takes into account the cross similarities between different pure and combined modes. The cross-nested structure allows for the introduction of the differentiated measurement of pairwise similarities among modes as opposed to the inflexible groupwise similarities permitted by the nested logit model. The proposed model is described, and it is compared with alternative modeling constructs.