Abstract
Popper famously held that the growth of scientific knowledge and the Darwinian mechanism of trial and error elimination are analogous processes. Both the validity of this analogy and Popper's interpretation of what this Darwinian mechanism consists in have been criticized. But it has been ignored that the use of Popper's Darwinian analogy had changed in the course of Popper's life. I will argue that until the 1960s, he used the Darwinian process as a model for understanding the growth of scientific knowledge, whereas from the 1960s on, the explanatory order was reversed: he used his new insights about the growth of scientific knowledge to say something about the real nature of Darwinian selection. In short, this analogy was so central for Popper's thinking that rather than giving up on it, he tried very hard to find theories of biological evolution that would make this analogy plausible. And this is what led him to make somewhat surprising claims about the nature of selection as well as to flirt with Lamarckism. I end by outlining a biologically plausible way of maintaining this Darwinian analogy that Popper failed to consider.