In both editions of his Theory of Knowledge (1990, 169–70 and 2000, 196), Keith Lehrer presents a case of a racist who becomes a racist medical doctor. Lehrer took the case to show (contrary to others’ arguments) that beliefs do not have to be based on the evidence in order to be justified, or count as knowledge. But there has been a mixture of responses to Lehrer’s thought experiment: not everyone who has considered the case has been ready to draw the same conclusions from it as Lehrer did. In this paper, I will give a kind of diagnosis of what may have led to resistance. I suspect that at least some of the resistance is the result of not accepting (or perhaps not understanding) the way in which Lehrer conceived of the case himself, because one essential ingredient of the case, as Lehrer conceived it, was not recognized or acknowledged by those who were left unconvinced.