This paper provides empirical evidence for the widespread belief that a magazine's companion website induces channel competition on its print version. The analysis is based on aggregated quarterly magazine-specific data on circulation, a broad variety of magazine characteristics, consumer characteristics and information on a magazine's website presence.The data spans the period I/1996 to IV/2004. Nested logit-type models of demand for differentiated products are used in the empirical analysis. A main result of this paper is that magazines that run an online companion on average lose 4.2 percent of their potential consumers. This effect varies substantially across different consumer age groups and across time.