Despite their prominence in state tax portfolios, state income taxes have received much less attention in the literature on behavioral responses to taxes. Consequently, decisions are often made without the needed evidence on response elasticities. In the spirit of the growing literature on the elasticity of taxable income, which has primarily considered federal taxes using individual data, we estimate state-level panel regressions of state personal income tax bases on state tax rates, structures, and other controls. Measuring state personal income tax bases is notoriously difficult, so we compare actual data gathered directly from state revenue officials with two other proxies: tax collections divided by the top marginal tax rate and total adjusted gross income on federal returns from each state. Results generally indicate that state tax rates on wage income have no impact on the actual tax base and only a small positive effect on the calculated base, but they have a larger effect on federal AGI. This suggests that individuals could be filing federal returns from higher-tax-rate states but are not necessarily reporting the same income to those states for state tax purposes. We also find that the tax rate on capital income is positively associated with state personal income tax bases. This might be picking up tastes for public services, to the extent that more mobile higher-income taxpayers move their entire tax bases to states with more or better services (and higher tax rates). Similarly, we find that the tax rate on pension income increases actual state base data while reducing the calculated base and federal AGI. The latter results might indicate that taxpayers with pension income prefer lower-tax-rate states and bring small tax bases with them when they relocate. It is important to note that, with the exception of tax rates on pension income, we find little to no evidence that higher state tax rates reduce personal income tax bases.