An orthodox characterization of the intra-household distribution of food suggests that men take 'unfairly' large shares, that this practice is the proximate cause of malnutrition and that women acquiesce in this distribution. This orthodoxy is regarded in the paper as a set of hypotheses, tested against the literature on food culture and behaviour, anthropometry and food intake. Food behaviour, set within a context of patriarchy, lends support to the orthodoxy qualified by diversity, both in women's structural positions and in types and conditions of household. The scope of existing nutritional data and concepts useful for evaluative purposes is critically reviewed. South Asian material reveals that discrimination against women in macronutrients tends to be a phenomenon restricted to North India and not always evident there. Discrimination against weanlings and 'pre-school' children of both sexes is more geographically widespread. If there is a pan-Indian problem, it is one of age rather than of gender. Poverty and/or scarcity do not seem to result in a distinctive alteration of intra-family shares. Gender discrimination is characteristic of distribution of certain of the micronutrients. Types of explanation for intra household maldistribution are reviewed and by way of conclusion it is suggested that the rise of intra household food distribution up the policy agenda is not the product of consistent research results and runs the risk of deflecting resources from improvements to food entitlements at the level of the household.