Human serine racemase (hSR) is a cytosolic pyridoxal-5′-phosphate dependent enzyme responsible for production of D-serine in the central nervous system. D-Serine acts as an endogenous coagonist of N-methyl-D-aspartate receptor ion channels. Increased levels of D-serine have been linked to amyotrophic lateral sclerosis and Alzheimer's disease, indicating that SR inhibitors might be useful tools for investigation or treatment of neurodegenerative diseases. However, despite hSR's promise as a therapeutic target, relatively few specific inhibitors have been identified, which is due in part to the lack of a three-dimensional structure of the enzyme. Here, we present a strategy for the generation and screening of random hSR mutants. From a library of randomly mutated hSR variants, twenty-seven soluble mutants were selected, expressed, and evaluated for enzymatic activity. Taking three carefully characterized mutants as an example, we show how this strategy can be used to pinpoint structurally and functionally important residues. In particular, we identify S84 and P111 as residues crucial for hSR activity and C217 and K221 as residues important for binding of the Mg 2+ cofactor as well as for overall stability of the enzyme. Keywords: D-Serine; Serine racemase; Error-prone PCR; Random mutagenesis; ThermoFluor assay; Enzymes; Racemases; Enzyme engineering; In vitro evolution.Eukaryotic serine racemase (SR; EC 5.1.1.18) is a cytosolic pyridoxal-5′-phosphate (PLP) dependent enzyme responsible for the production of D-serine from L-serine and vice versa. In the mammalian central nervous system (CNS), D-serine acts as an endogenous coagonist of N-methyl-D-aspartate (NMDA) receptors, which are involved in glutamate-mediated excitatory neurotrans-