This article reviews the issues involved in policy choices with respect to universality and selectivity in income support to older people. It considers four questions: the practical meaning of universality and selectivity in the income support systems of various countries, the effectiveness of universal and selective arrangements in the alleviation of poverty among this group, the role of universal and selective arrangements in redistributing income among elderly people and the relative generosity of universal and selective arrangements. The article draws on data from the ‘second wave’ of the Luxembourg Income Study for six countries: Australia, (West) Germany, Norway, Sweden, the United Kingdom and the United States, concerning the incomes of elderly couples and single (non-married) women. It concludes that while selective income support arrangements achieve greater redistribution in favour of low income elderly people for the same expenditure than do universal ones, selective arrangements do not necessarily perform better in other respects, and, in particular, are associated with low levels of benefit income.