This study compares the diet of Greek Neolithic populations to determine the relationship between nutrition and pathology during the transition to agricultural economy. Stable isotope analysis of carbon and nitrogen from human bone collagen and carbonate apatite was used on 102 samples representing six populations, three coastal, (Alepotrypa, Franchthi, Kephala), and three inland, (Theopetra, Tharrounia, Kouveleiki). The results, namely the very negative carbon values, the low nitrogen values, and the intermediate values of the difference between carbonate and collagen carbon readings, strongly indicate the consumption of a diet focused on C3 terrestrial resources, including mainly agricultural plants and almost certainly some animals. Marine foods were a minor part of the diet of all sites. Furthermore, the small range of values implies homogeneity in their diet. The analysis suggests that Neolithic Greece, regardless of geographic location, was occupied by agricultural groups with primarily terrestrial subsistence strategy with only occasional exploitation of animal and marine protein resources. Nutritionally, the combination of cereals and pulses is a rich source of energy, protein and amino-acids, but does not supply all the required vitamins and iron. Based on the aforementioned data an attempt was made to determine the impact of such a diet on human health. The most prevalent pathological conditions among Greek Neolithic populations are mild cribra orbitalia and porotic hyperostosis, anaemic conditions resulting from increased production of red blood cells, in response to different kinds of anaemia of genetic or nutritional origin or/and in combination with bacterial and parasitic diseases. Even though this study does not provide a direct association between health and diet, it corroborates evidence that diet cannot be excluded as a contributing factor to the observed anaemic conditions, which can possibly be attributed to iron deficiency anaemia. However, hereditary anaemia is excluded by the pathological manifestation of the lesions, while parasitism and bacterial load can interfere and, therefore, it must be checked if they are contributing factors as well. Copyright © 2003 John Wiley & Sons, Ltd.