Abstract: The Relevant Alternative (RA) approach to the semantics of knowledge ascriptions blends epistemic fallibilism with epistemic modesty. Schaffer [2002] introduces a family of alleged counter-examples to this approach: missed clue cases. These have received limited attention in the literature: agreement was quickly reached that they only arise for RA theories that do not properly integrate a belief condition. I argue that the significance of missed clue cases has been under-estimated: their import has nothing to do with the interaction between belief and knowledge. Rather, an RA theory is subject to them exactly when that theory violates a particular epistemic closure condition. The RA theorist has two strategies in response: maintain this particular closure condition, or motivate its violation (in line with violations of similar closure conditions). I defend the second strategy.