We examine the historical dynamics of government debt in post-unification Italy, from 1861 to 2009. Unit root tests for the debt-GDP ratio are unable to reject either the non-stationarity or the stationarity null hypothesis. Controlling debt dynamics for fiscal feedback policies of the Barro-Bohn style, however, the debt-GDP ratio is found to be mean-reverting. Mean-reversion in the debt-GDP ratio is due not only to a nominal growth dividend, but also to a positive response of primary surpluses to variations in outstanding debt. There is indeed significant evidence that, over the history of Italy, fiscal policy makers have reacted to the accumulation of debt, taking corrective measures to rule out potential long-term sustainability problems.