How do people decide which claims should be considered mere beliefs and which count as knowledge? Although little is known about how people attribute knowledge to others, philosophical debate about the nature of knowledge may provide a starting point. Traditionally, a belief that is both true and justified was thought to constitute knowledge. However, philosophers now agree that this account is inadequate, due largely to a class of counterexamples (termed "Gettier cases") in which a person's justified belief is true, but only due to luck. We report four experiments examining the effect of truth, justification, and "Gettiering" on people's knowledge attributions. These experiments show that: (1) people attribute knowledge to others only when their beliefs are both true and justified; (2) in contrast to contemporary philosophers, people also attribute knowledge to others in Gettier situations; and (3) knowledge is not attributed in one class of Gettier cases, but only because the agent's belief is based on "apparent" evidence. These findings suggest that the lay concept of knowledge is roughly consistent with the traditional account of knowledge as justified true belief, and also point to a major difference between the epistemic intuitions of laypeople and those of philosophers. (Contains 4 figures.)