I argue that knowledge-first epistemology has significant consequences for the philosophy of science.  In particular, the identification of evidence with knowledge undermines empiricist philosophy of science and refutes the Duhem–Quine thesis, while supporting a modest kind of scientific realism.