There was a time, and not so very long ago, when research in the geography of agriculture had little to say about food other than as a raw commodity. By the same token, geographers of food were largely taken up with retail geography as a subspecialism within economic geography. In other words, both were about economic activity, but in separate compartments. Both focused on issues of production as against consumption. The emergence of an agro-food geography that seeks to examine issues along the food chain or within systems of food provision derives, in part, from the strengthening of political economy approaches in the 1980s. By directing attention away from the narrow confines of the farm business, hitherto examined largely through the lenses of either neoclassical economics or behaviourism, towards relations with ‘capital’ an important reorientation and reconnection was started. However, the process has been taken much further in recent years with the (re-)discovery of consumption, linked so strongly to the postmodern or poststructuralist trends in social science. Conceptualizing consumption is, in turn, linked to the discovery of ‘culture’ in economic geography, and this has had some positive impacts on agro-food studies (Goodman and DuPuis, 2002). However, Martin and Sunley (2001) have recently raised concern that ‘what these cultural economic geographers criticise as the myopic economism of ‘‘old’’ economic geography could simply be replaced by an exclusionary cultural essentialism in the ‘‘new’’ ’ (p. 152). To date, cultural essentialism has not taken hold in agro-food studies where the legacy of political economy is strong (e.g., Marsden et al., 1993). Consequently, economic realities and power remain important in many discussions of food consumption, as in Marsden et al. (2000). Nonetheless, a negative impact of the new cultural geography of consumption is perhaps apparent in how very few recent studies highlight the basic socio-economic and political issues of income inequalities which underpin and result from differential access to the process and Progress in Human Geography 27,4 (2003) pp. 505–513