When the issue of welfare is addressed in Sweden analysis is usually focused on the national level, in a smaller number of cases on the local level, but rarely on the regional level. Studies on poverty are no exception. The aim of this paper is to analyse child poverty in Sweden, from a regional perspective, in 1990 and 2010. A multivariate cross-section OLS regression model is used to estimate regional factors causing child poverty. Regional macro data is used. The findings indicate that old industrial regions experience greater child poverty. The statistical analysis shows that when the share of foreign-born individuals increases in the regional population, so does the regional share of children living on social assistance. Higher incomes lower the regional share of children living on social assistance.