This paper applies tax incidence theory to estimate the distributional effects of the exemption from federal income tax of interest on state and local bonds and the President's proposal to limit the benefit of the exemption to the 28 percent rate. When one accounts for the effects of changes in rates of return, the exemption still primarily benefits higher-income individuals, though less so than under the traditional approach that assigns all the benefit to holders of tax-exempt bonds. How states and localities respond to lower borrowing costs can either shift benefits to low-income households or increase the net benefit to high income households.