The combination of protein design and directed evolution is a powerful way to create enzymes with new functions. Now, Anthony P. Green and coworkers at the University of Manchester have thrown a nonnatural amino acid into the mix. They started with a protein computationally designed to catalyze Morita-Baylis-Hillman carbon-carbon bond formation and transformed it into an enzyme that catalyzes ester hydrolysis (Nature 2019, DOI: 10.1038/s41586-019-1262-8). First, they replaced the catalytic histidine in the active site with a noncanonical methylhistidine. Then they subjected that new enzyme to multiple rounds of laboratory evolution with an assay based on fluorescein detection. The most active variant hydrolyzed fluorescein 2-phenylacetate 9,000 times as efficiently as free methylhistidine in solution and 2,800 times as efficiently as the organocatalysts dimethylaminopyridine and N-methyl imidazole. Further rounds of evolution gave rise to an enantioselective enzyme that preferentially hydrolyzes one of the two mirror-image isomers of fluorescein 2-phenylpropanoate. Modifying the methylhistidine completely shut down the evolved enzymes, confirming it as the key catalytic amino acid. The ability to include nonnatural catalytic amino acids could expand the range of chemistries available in lab-evolved enzymes.