Background: ecological studies using standard generic indicators of material deprivation have suggested that there is little social inequality at older ages. This may be because the indicators used were designed for studying younger populations and may be biased against older people. Objective: to examine the association, at different ages, between mortality ratios, indicators of deprivation and an indicator of poverty, which can be easily tailored to different age groups. Methods: an ecological study comparing mortality ratios and indicators of deprivation and poverty. We calculated standardised mortality ratios for those under and over 75 years, using all deaths for Northern Ireland between 1990 and 1998. We calculated levels of income support uptake (a social security benefit) for similar age groups. We derived three commonly used indicators of deprivation (Townsend, Carstairs and Jarman) from census data. We assessed the strength of association between the mortality ratios and the indicators of poverty and deprivation using Pearson correlation coefficients. Results: 11.1% of people under 75 and 24.3% of those over 75 were on income support. Income support uptake for those over 75 was strongly correlated with deprivation indicators. There was a much weaker relationship between disadvantage and mortality at older ages, especially in women. Poverty, as measured by income support ratios, was more highly correlated with mortality than any other deprivation indicator—especially at older ages and in women, where income support uptake produced the highest correlations. Conclusions: many of the commonly used indicators of deprivation are poorly suited to studying inequalities in health in older people. Uptake of income support offers many advantages over conventional indicators.