This study evaluates how major changes in Uzbekistan’s personal income tax (PIT) and value-added tax (VAT) regimes altered the distributional structure of taxation during 2015–2025. The analysis combines official income-decile distributions, Gini coefficients, household income composition, household final consumption expenditure and spending composition with historically coded PIT and VAT rules. It applies the established Lorenz– concentration-curve framework of Kakwani (1977), complemented by effective tax-burden analysis and the Reynolds–Smolensky logic of redistribution. Because complete annual household tax-payment microdata is not publicly available for the entire period, the paper is explicitly designed as a real-data-anchored longitudinal microsimulation: official statistics and legal tax schedules are observed inputs, whereas VAT/excise incidence and some PIT distributional quantities are modeled. The progressive PIT schedules of 2015–2018 generate positive simulated Kakwani indices of 0.1535–0.1778. Under the structural benchmark in which the 12% post-2019 PIT is proportional to the ranking income base; PIT progressivity falls to zero. Modeled VAT remains moderately regressive throughout the period, with Kakwani indices near −0.05, while modeled excise incidence is more regressive. Reductions in the VAT rate from 20% to 15% in 2020 and to 12% in 2023 lower the level of indirect taxation but do not mechanically remove its relative regressivity. The results imply a shift from a tax mix containing a strong progressive direct-tax component toward one in which that offset is substantially weaker. The paper therefore distinguishes changes in statutory rates, progressivity, tax levels and redistributive effects, and cautions against interpreting before/after movements as causal effects without household-level counterfactual identification.