Recent research indicates that political polarization in Congress and income inequality share a closely linked positive association. But virtually no studies examine the direction of influence between these variables as it is assumed that income inequality causes political polarization. The major purpose of this investigation is to examine the temporal causal ordering of these variables. Methods. This study constructs a time series national‐level data set with information for the years 1913 to 2008. Vector autoregression and granger causality tests are utilized to explore the temporal causal ordering of congressional polarization and the income share of the top 0.1, 1.0, 5.0, and 10.0 percent of earners in the United States. Autoregressive conditional heteroskedasticity regressions are also employed to assess the strength of the association between congressional polarization and top incomes net of relevant control variables. Results. The findings indicate that the past values of congressional polarization are better predictors of top income shares than vice versa. The results also demonstrate that polarization in the House of Representatives produces a more consistent and robust connection with top incomes than polarization in the Senate. Lastly, congressional polarization only produces robust associations with the income share of the top 0.1 and 1.0 percent of earners but not for the top 5.0 and 10.0 percent. Conclusion. While the Senate possesses more powerful negative agenda control procedures to stifle the legislative processes vis‐à‐vis the House, it is polarization in the latter that returns the more robust associations with income inequality.