In the light of violent clashes between Kyrgyz and ethnic Uzbeks in Kyrgyzstan in June 2010 we investigate the association between economic welfare and ethnicity in this country. We intend to answer two questions. First, are Uzbek households better off than Kyrgyz households, as is often claimed in the media and also by some academics? Second, what are the correlates of household welfare in recent years, and how have these changed in comparison with the 1990s? We use data from two cross-sections of the Kyrgyz Integrated Household Survey (2003 and 2005) and run OLS regressions using three measures of welfare, i.e. per capita consumption, per capita income, and an asset index. We find some evidence for higher welfare of Uzbek headed households compared with their Kyrgyz counterparts, but mainly in rural areas. In the south of the country, where most Uzbeks live and where the violence took place, there appears to be no substantial difference in welfare. This is clearly in contrast to what was commonly propagated in the media and what most Kyrgyz tend to think. In terms of the other correlates of welfare, we find that household size, educational attainment of adults, and residence outside the capital and the neighbouring Chui oblast are most importantly connected with welfare. This coincides with findings from earlier studies using data from a decade earlier.