4-Amino-2-(substituted methyl)-2-butenoic acids, where X (the substituted group) = F, Cl, OH, are synthesized from Cbz-protected tert-butyl 4-aminobutanoate. Successive substitutions at the alpha-carbon by phenylseleno and hydroxymethyl groups, followed by elimination of the selenoxide and halide substitution at the hydroxymethyl group, afford the compounds in good yields. An unexpected degree of stereoselectivity is observed in the selenoxide elimination step, which yields the desired E isomer as the sole product. These compounds complement two previously reported series of compounds (Silverman, R. B.; Levy, M. A. Biochem. Biophys. Res. Commun. 1980, 95, 250-255; J. Biol. Chem. 1981, 256, 11 565-11 568) and are used in an approach to map a section of the active site of gamma-aminobutyric acid aminotransferase (GABA-T). None of these compounds is a time-dependent inactivator of GABA-T, but all are potent competitive reversible inhibitors; the hydroxy compound has a Ki value of 5 microM. That these compounds are not inactivators suggests that either elimination of X does not occur or that there is no active site nucleophile in the appropriate position for reaction following elimination. With use of the fluoro analogue, enzyme-catalyzed fluoride ion release is demonstrated, indicating that elimination does occur. Unlike the previous two series of compounds (op. cit.) in which exclusive elimination occurs when the substituent is a halogen but exclusive transamination prevails for the hydroxyl-substituted analogues, in the series described here, the fluoro analogue gives a 4:1 ratio of elimination to transamination. This suggests that the 2,3-double bond stabilizes the product of azallylic isomerization of the Schiff base between the fluoro compound and pyridoxal phosphate. The results described here indicate that the design of a mechanism-based inactivator for GABA-T should not be based on electrophile generation near the 2-position of enzyme-bound GABA. Furthermore, substitution of an inhibitor with a 2-hydroxymethyl group (or other hydrogen-bonding substituent) and a 2,3-double bond may lend auspicious binding properties to the molecule for GABA-T.