We study the association between natural resource rents and internal political stability, highlighting the importance of the distribution of political power as a mediating factor. We present a simple theoretical model demonstrating that increased rents are likely to be positively associated with the internal stability of a powerful incumbent while destabilizing a less powerful incumbent. Our empirical analysis confirms this prediction. Employing panel data for more than 120 countries from the period 1984–2009, our estimation results demonstrate that resource rents can promote political stability but only when political power is sufficiently concentrated. Indeed, if the incumbent is sufficiently weak, rents fuel instability. Our main results hold when we control for the effects of income, quality of institutions (rule of law, democratic accountability and corruption), persistence of political stability, time-varying common shocks, country fixed effects, possible endogeneity of rents and power balance to political stability, and various additional covariates. Our analysis departs from the existing literature by emphasizing not (only) the type of government, but rather the strength of government, as a key determinant of the impact of resource rents on political stability. This analysis sheds light on current political transformations and reconfigurations in the resource rich Middle East and North Africa.