In this paper, we employ stochastic dominance analysis on Chinese Household Nutrition Survey (CHNS) data to investigate the inequality and relative welfare levels in China over time and among population subgroups. We find that from the period of 2000 to 2009, welfare has been continuously improved along with Chinese economic development and growth. Our pairwise comparison of population subgroups shows that there is no dominance relation between subgroups for household type, gender of households head, and age cohorts. While married group and non-child rearing group second order dominate single/divorced group and child rearing group, showing higher level of welfare in the former groups. Also, we find inequality in subgroups with different educational levels and household sizes that the groups with a higher level of education and smaller size of household tend to be better off than their counterparts.