How do sovereign debt restructurings reduce debt-to-GDP ratios? We explore this empirically using a comprehensive dataset covering 115 countries over 1950–2021. After addressing selection bias through an Augmented Inverse Probability Weighted estimator, we show that restructurings significantly reduce debt-to-GDP ratios over 1-5 years, with the effects working primarily through debt levels. The effect is larger when restructurings are combined with fiscal consolidation. We find heterogeneity depending on the creditor type, and the type and size of debt relief. In the short run, restructurings with higher creditor coordination, face value reductions, and larger debt reliefs, reduce debt-to-GDP ratios more effectively. (JEL F34, F41, H63)