It is well known that Uganda has achieved impressive progress in poverty reduction over the 1990s, based to a large extent on a good macroeconomic performance in combination with a specific package of poverty eradication measures. Monetary poverty fell from 56% of the population in 1992/93 to 35% in 1999/2000. Large numbers of households escaped from poverty, but equally a substantial number of households remained in persistent poverty throughout this period. Such chronic poverty typically reflects particularly deep-seated disadvantages, so tackling this poverty is likely to be an especially difficult challenge over the next years as the government seeks to approach its poverty eradication aim. This paper focuses on the extent and nature of chronic poverty in Uganda over this period, and the likely implications for policy, in particular the Poverty Eradication Action (PEAP). It is based particularly on the good quality household panel data set available for up to four rounds between 1992 and 1996, though also draws on evidence from the 1992 to 1999 panel data set. The results show that chronic poverty in Uganda is not only location-specific but depends on various initial household characteristics. The findings have important policy implications, in that the chronically poor appear not to have benefited much from the market-oriented development policies that have been responsible for much of Uganda's macroeconomic success over this period.