A series of enantiomerically pure (phosphonomethyl)-substituted phenylalanine derivatives related to SDZ EAB 515 (1) were prepared as competitive N-methyl-D-aspartate (NMDA) receptor antagonists. Unlike most known competitive NMDA antagonists, analogs in this series with the S-configuration are potent NMDA antagonists whereas analogs with the unnatural R-configuration are weak NMDA antagonists, as determined by receptor binding experiments and their anticonvulsant action in mice. Examination in a previously reported competitive NMDA pharmacophore model revealed that receptor affinity can be explained partially by a cavity that accommodates the biphenyl ring of 1, while the biphenyl ring of the R-enantiomer 2 extends into a disallowed steric region. We proposed that analogs with the natural S-configuration and a large hydrophobic moiety would have an advantage in vivo over analogs with an R-configuration by being able to use a neutral amino acid uptake system to enhance both peripheral adsorption and transport into the brain. Examination in a system L neutral amino acid transport carrier assay shows that 1 competes with L-Phe for transport in an apparent competitive and stereospecific manner (estimated Ki = 50 microM). The 1- and 2-naphthyl derivatives 3a,3b were found to be among the most potent, competitive NMDA antagonists yet discovered, being ca. 15-fold more potent than 1 in vitro and in vivo, with a long duration of action. The title compound 3a had potent oral activity in MES (ED50 = 5.0 mg/kg). 3a also retains its ability to compete, albeit more weakly than 1 (estimated Ki = 200 microM), for L-Phe uptake to CHO cells. In this series, analogs with the R-configuration are not substrates for the system L neutral amino acid transport carrier. These results provide evidence that central nervous system active agents can be designed as substrates of a neutral amino acid transporter as a means to enhance penetration of the blood-brain barrier.