Intermediation in markets is a phenomenon that has been studied by many researchers from a variety of different theoretical angles. With the introduction and diffusion of the Internet in everyday life, broad predictions were made that called for disintermediation enabled by direct Internet linkages between suppliers and buyers and lower transaction costs. The often-cited paper by Sarkar, Butler and Steinfield (1995) challenges this prediction. By comparing Internet effects on transaction costs with the cost situation ex ante, the paper explains that both direct sales or cybermediated sales are possible outcomes. In this paper we confront key assumptions of the Sarkar et al. paper with recent developments in the tourism market. We find that in the tourism market a multitude of direct and indirect distribution channels exist next to each other. Multi-level distribution channels often including several cybermediaries have been built, resulting in a complex market topology. We also see a large variety of intermediary roles, resulting from highly specialized and highly integrated cybermediary business models. Furthermore the model of Sarkar et al. fails to deliver an explanation for the on-going dynamics in the tourism market in terms of shifts towards more or less intermediaries and the emergence of new intermediary-like business models. By taking these trends into account we are able to identify relevant future research directions in order to extend our understanding of the phenomenon of electronic intermediaries in markets.