Arguments from ignorance are typically fallacious: one shouldn’t infer that p is true merely from the fact that p isn’t known to be false. The reason one cannot invariably come to know that p on the basis of lack of knowledge that not-p is that p may well be false even though one hasn’t been in a position to know that not-p. The qualifiers ‘typically’ and ‘invariably’ are key here. In cases where one would have known that not-p had p been false, reasoning from ignorance seems perfectly kosher, indeed where one knows that counterfactual independently, one is positioned to gain knowledge. Goldberg (2010a; 2010b) has developed a set of conditions sufficient for such epistemic coverage to obtain in the domain of social epistemology. This paper elaborates, extends and critically discusses Goldberg’s pioneering framework.