Premise 2 of the argument by counterexample claims that transmission fails in Dretske cases: S can’t acquire a warrant for Q by inference from P, even if she is warranted in believing P. To deny this is to claim that transmission always succeeds in Dretske cases. I argue that it doesn’t. After pointing out that it is unintuitive that transmission does succeed in such cases, I show that transmission failure also follows from a highly intuitive general principle limiting warrant acquisition: if the putative method of warrant-acquisition itself ensures that, whenever a proposition R is true, that method will inevitably deliver the result that R is false, then that method cannot deliver a warrant for the claim that R is false. I call this the “NIFN” – no inevitable false negatives – rule. The zebra case violates NIFN, which implies that transmission fails in that case. I distinguish NIFN from appeal to sensitivity, indicating that standard objections to sensitivity as a condition of warrant don’t undermine appeal to NIFN. I then defend NIFN against objections based on the generality problem and method externalism, and generalize the application of NIFN to other Dretske cases.