The study of the proportions of stable isotopes of carbon and nitrogen which survive in ancient human and animal bones offers highly suggestive indications of ancient diets. Among the most remarkable results from such investigations is the dramatic change in diet which is thought to have occurred between the Mesolithic and the Neolithic when people turned from maritime to terrestrial food, from fish to meat and vegetables. The three contributions which follow challenge, modify, enhance or reflect on this model. In a pivotal critique of the evidence from Britain and Denmark, Milner et al. present a range of explanations for the signals of a maritime or terrestrial emphasis in diet and conclude that the change need not have been either rapid or total. Lidén et al. show that, in southern Sweden, the preferences for fish over meat were related less to period or culture, but (reasonably enough) to location: fish-eaters live by the sea. Finally Robert Hedges takes up the question of partial marine diets and how to detect them, developing the idea that marine diets might give a fainter signal in people who were only getting small amounts of protein. Perhaps there were many such people in the new order of the Neolithic …