Can autonomous banking regulatory agencies reduce the odds that a country will suffer a crippling banking crisis? We investigate the impact that agencies charged with banking regulation and prudential supervision can have on financial stability in the banking sector. We argue that the potential benefits of autonomy are hard to realise because banking regulators face incentives to shirk in their mandate to secure banking stability. These incentives are strongest in political systems with high numbers of veto players, where the autonomy of a banking agency is difficult to undo even if the agency is derelict in promoting banking sector stability. We test an implication of this argument, namely, that the probability of bank crisis onset should diminish with the level of autonomy of the banking agency, but only in polities with low numbers of veto points. We base our analysis of this conditional hypothesis on an original dataset of 79 countries observed between 1971 and 2009 that captures the degree of autonomy of banking agencies from political principals. Our findings confirm that the impact of banking agency autonomy on the risk of bank crisis onset is conditional on the political structure in which the agency is embedded.