Important reactions in biology and biocatalysis involve proton abstraction from carbon. When the resulting anionic charge is delocalized from carbon to an oxygen atom, these deprotonations can be catalytically accelerated by oxyanion stabilization. Oxyanion stabilization by a glutamine side chain (Gln50) was thought to accelerate C–H proton abstraction in HG3.17, a computationally designed biocatalyst that had been evolutionarily optimized to enzyme-like efficiency. We present kinetic data and crystal structures at atomic resolution for six Gln50 mutants that indicate a surprisingly small advantage of the hydrogen-bond donor glutamine over “greasy” methionine. However, tightly packed active sites (Gln, Met, Phe)—greasy or not—perform consistently better than water-filled oxyanion holes found with other substitutions (His, Ser, Ala, Lys). Although oxyanion stabilization appears to contribute modestly to HG3.17 efficiency, the role of Gln50 is mechanistically more complex than initially thought, underscoring the importance of multifactorial approaches for the design of enzymatic oxyanion holes in the future.