Many of the mutations that are accumulated by naturally evolving proteins are neutral in the sense that they do not substantially affect a protein's ability to perform its primary biological function. Here I describe experimental work with cytochrome P450 BM3 enzymes demonstrating how these initially neutral mutations can nonetheless have profound effects on later functional evolution. I discuss two mechanisms by which initially neutral mutations can enable future functional evolution: by modulating an enzyme's promiscuous activities, or by altering its stability in a way that changes it tolerance for future mutations. I will also show that the changes in promiscuous activities of cytochrome P450s that have undergone neutral evolution can often be rationalized in terms of the chemical structures of the enzymatic substrates using the computational technique of principal component analysis. Finally, I will briefly touch on the application of these findings in designing more effective strategies for the engineering of enzymes.