One prominent argument for pragmatic encroachment (PE) is that PE is entailed by a combination of a principle that states that knowledge warrants proper practical reasoning, and judgments that it is more difficult to reason well when the stakes go up. I argue here that this argument is unsuccessful. One problem is that empirical tests concerning knowledge judgments in high-stakes situations only sometimes exhibit the result predicted by PE. I argue here that those judgments that appear to support PE are better interpreted not as judgments that the epistemic demands for knowing increase as one’s practical situation becomes more demanding, but instead as judgments reflecting a different kind of normative epistemic evaluation, namely whether one is acting in an epistemically responsible way. The general idea is that when someone treats a proposition as a reason for acting we can evaluate them epistemically both in terms of whether they know that proposition, as well as in terms of whether they are acting on their knowledge in the right kind of way. My charge against the PE proponent, then, is that she is interpreting judgments that are indicative of whether we are adhering to certain normative epistemic requirements generally as being indicative of whether we have knowledge specifically. There are, however, normative epistemic requirements that make demands of us that are indicative of something other than our possession of knowledge.