The philosophical foundations of Hayek’s works are not beyond dispute (Gray, 1984, Kukathas, 1989, Caldwell, 1992, Hutchison, 1992): was Hayek a rationalist or an empiricist; did he follow Kant or Hume, Mises or Popper? Difficulties arise because these questions touch upon social theory, political philosophy, methodology and epistemology. Moreover, on different occasions, Hayek (intentionally) gave different definitions and evaluations of already complicated views such as ‘rationalism’ and ‘empiricism’. In this paper, I try to shed some light on the rationalism/empiricism issue by focusing on epistemology, where this issue really belongs. The debate there is mainly about the sources of knowledge (e.g., Markie, 2008). Empiricists argue that experience is the source of all our knowledge. This view was held by John Locke (1632-1704) and David Hume (1711-1776) but its roots go back to Francis Bacon (1561-1626) and even further to the ancient Greek Empiricist school in medicine (founded in the third century B.C. by Philinos of Kos or Serapion of Alexandria) and Aristotle (384-322 B.C.). In contrast with his teacher Plato, Aristotle believed in the ‘induction’ (epagōgē) of general knowledge from particular observations. I will not have the space here to relate Hayek’s ideas to this long history of empiricism. But I will try to refer to David Hume now and then, because Hayek was a great admirer of Hume’s social and political philosophy and Hayek’s ‘Humeanism’ is extensively discussed. I will also get back to the less well known Empiricist school in medicine, because it has a very special conception of ‘experience’ which I believe to be useful to the discussion. In contrast with empiricism, rationalism or ‘apriorism’ is the idea that some knowledge is independent of experience or ‘a priori’. Traditionally, this meant that knowledge is based on rational intuition, or embedded in our rational nature or the structure of the mind. If knowledge is embedded in our mind or nature, it is ‘innate’, which is why philosophers speak of ‘innatism’ or ‘nativism’. Since this was Immanuel Kant’s (1724-1804) view, it is often called ‘Kantianism’. I will also use the term ‘Kantianism’ rather than ‘rationalism’ because Hayek most often defines the latter as the false view that social phenomena are rationally designed, which is a completely different issue. Kantianism goes back to the ‘innate ideas’ of René Descartes (1596-1650) and the anamnesis of ideas in Plato’s philosophy (429-347 B.C.). Many scholars have tried to position Hayek in the Kantianism/empiricism debate. Most scholars would probably agree with Connin (1990, p. 301) that “Hayek’s theory of knowledge is undoubtedly Kantian” (see also Feser, 2006, p. 300). However, many also understand that there is more to it (Caldwell, 2004, p. 273). Since ‘experience’ is undeniably a basic concept in Hayek’s epistemology, some believe that his epistemology is a kind of synthesis between Kantianism and Humean empiricism (Horwitz, 2000, p. 25). De Vecchi (2003, p. 152) is less optimistic and says that “there is an unresolved tension between empiricism and anti-empiricism within the theory of the process of the formation of knowledge set out in The Sensory Order”. Moreover, some have made the link with ‘evolutionary epistemology’ (Bartley, 1987, p. 21; Gray, 1984; Kukathas, 1989; Dempsey, 1996; Vanberg, 2002). However, scholars have rarely wondered how Kantianism, empiricism and evolutionism can be reconciled, and, more importantly, what ‘empiricism’ and ‘experience’ mean in such a context. Just as there are as many ‘rationalisms’ as there are interpretations of the term ‘reason’, there are as many ‘empiricisms’ as there are interpretations of the term ‘experience’. In this paper, I will reconstruct Hayek’s epistemology based on a careful reading of The Sensory Order and some related writings. I will argue that Hayek’s epistemology is best characterized as a type of ‘post-positivist empiricism’. In the first paragraph, I review Hayek’s neurophysiological explanation of the mind in The Sensory Order. Hayek shows how the nervous system can perform the acts of classification characteristic of the working of the mind. Because the synaptic connections embody a kind of knowledge independent of ‘sense experience’, Hayek is not a ‘sensationalist empiricist’. The second paragraph discusses Hayek’s theory of the formation of synaptic connections. Connections are formed on the basis of what I will call ‘Hayek’s learning rule’, which boils down to the familiar idea that neurons that fire together wire together. Since this means that the knowledge embodied in the synaptic connections is in a sense the result of ‘experience’, be it ‘pre-sensory experience’ rather than ‘sense experience’, Hayek is an empiricist after all, but one of the ‘post-positivist’ kind. In the third paragraph, I analyze Hayek’s views on the evolution of the nervous system and the behavior it generates. There appear to be two kinds of ‘experience’ at the basis of the synaptic connections: ‘experience of the individual’ and ‘experience of the race’. Because Hayek denies that all knowledge is due to ‘experience of the individual’, he is not an ‘individualist empiricist’. However, since ‘experience of the race’ is also ‘experience’, he is again an empiricist in the wider sense. What Hayek failed to notice is that experience of the race is ‘post-sensory’ rather than ‘pre-sensory’ and also in other aspects very different from individual experience. I will call it a kind of ‘selective experience’, which I contrast with ‘inductive experience’. Some links with Donald Campbell’s ‘evolutionary epistemology’ are explored. In the last paragraph, I consider Campbell’s idea that all increases in knowledge are due to selection and make some suggestions for future research.