This paper presents new evidence on how top income earners respond to changes in the personal labor income tax schedule, uncovering both own- and cross-tax base responses within a unified framework. For identification, we exploit a 2012 tax reform in Uruguay that generated quasi-random variation in top marginal rates within the top 1% of the labor income distribution. Our empirical approach relies on a difference-in-differences identification strategy and administrative records linked at the individual level across multiple tax bases. We estimate an own-tax base intensive margin elasticity of 0.82 and extensive margin semi-elasticity of 3.36. Extensive margin responses are mostly driven by taxpayers shifting from the personal labor income tax base toward corporate income or capital income tax bases (semi-elasticities of -1.00 and -0.61, respectively). Our preferred estimates suggest that the reform was effective in increasing tax revenues, with efficiency costs representing 31% of the projected increase. However, it had limited impact on inequality, most likely due to its narrow scope and income shifting toward tax bases with lower and flat rates. Overall, our results indicate that policy efforts aiming to reduce inequality by increasing top marginal tax rates should also focus on limiting income shifting opportunities to strengthen their redistributive effects.