As developing countries attempt to improve public services and develop redistributive transfer policies, there is a pressing need to increase government revenues, through both direct and indirect taxation. As part of deciding which taxes to increase, one must study the distributional impact of tax reform and the implied welfare changes, as well as consider the implications of policy for economic efficiency. This presentation will detail a tax microsimulation tool developed for Mexico (MEXTAX), and the methodology used in applying this to policy questions. An application to the 2010 reforms to income tax, VAT and excise duties will be discussed and presented. In the context of poor micro-data, we emphasise the importance of testing the sensitivity of results to the assumptions one makes about under-recorded income and tax evasion. We also investigate how incorporating behavioral response along several margins may affect results. Together these robustness checks suggest a need for significant improvements to the micro-data available and in modelling how changes in tax rates affect taxable income in Mexico.