Recent work has documented a rising degree of wealth inequality in the United States between 1983 and 1998. In this work, we look at another dimension of the distribution, polarization. Using techniques developed by Esteban and Ray (1994) and further extend by D'Ambrosio (2000), we examine whether a similar pattern exists with regard to trends in wealth polarization over this period. The approach here followed provides a decomposition method which allows one to monitor what factors modified the entire distribution, where precisely on the distribution these factors had an effect and what determined the variation in the level of social distance between groups. These effects are estimated by applying kernel density methods to appropriate samples in order to obtain counterfactual distributions. An index of social distance is provided as well as summary statistics of the observed movements and of distance and divergence among the estimated and the counterfactual distributions. The decomposition method is applied to USA data on the distribution of wealth between 1983 and 1998. We examine polarization patterns and their change over time with regard to a number of household dimensions: home owners and renters, race and ethnicity, age class, family type, household income class and educational class.