Since the emergence of early endogenous growth models (Larry E. Jones and Rodolfo E. Manuelli, 1990; Sergio Rebelo, 1991), a large body of work has studied the growth effects of tax reform. While virtually all of this literature has confined itself to the analysis of flat-rate taxes, tax codes are generally progressive. This paper, therefore, explores the effects of progressive taxes in conventional growth models with heterogenous households. In such frameworks, the tax code helps to determine simultaneously the pre-tax income distribution and the rate of technical progress. We present three main findings. First, we show that when variations in tax codes stem from differences in progressivity, shares of tax revenue in GDP or income may constitute poor proxies for average marginal tax rates. In particular, we find that the decrease in progressivity associated with the Tax Reform Act of 1986 (TRA-86) lowered the U.S. average marginal tax rate by 0.06 to 0.37 percentage points depending on the model used. At the same time, however, the endogenous adjustment in the distribution of income produced by this progressivity change contributed to raising the tax share of income by approximately 0.8 percentage points. Because marginal tax rates are not easily observable, empirical work often has to make use of tax shares in income as an alternative. To the degree that lower marginal rates entail less tax distortion, our study suggests that relying on tax shares may cause less progressive tax codes to be incorrectly perceived as more distortional. Second, we find that the progressivity decrease implied by TRA-86 helped raise U.S. per capita GDP growth by 0.12 to 0.34 percentage points. Given the prominence of TRA-86 compared with other U.S. tax reforms over the past four decades, these growth effects, while not negligible (as in Robert E. Lucas, 1990), shed doubt on the potential for tax policy to alter significantly prospects for U.S. long-run growth. Finally, consistent with previous work, our analysis suggests that the progressivity change associated with TRA-86 had a significant effect on income inequality, resulting in a 20to 24percent increase in the Gini coefficient of income. We carry out our analysis using two prototypical endogenous growth models augmented to include a nondegenerate distribution of income. These models, one first formulated by Robert J. Barro (1990) and the other by Rebelo (1991), account for two polar extremes regarding the use of tax proceeds. On the one hand, in Rebelo’s two-sector framework, tax revenue is spent in a way that affects neither the marginal utility of private consumption nor the production possibilities of the private sector. On the other hand, in the environment envisioned by Barro (1990), all tax revenue serves to finance public services that enhance private production. Interestingly, the results we have just described emerge irrespective of the framework under consideration.