A novel series of octahydrophenanthrenamines and their heterocyclic analogues have been synthesized as potential noncompetitive antagonists of the N-methyl-D-aspartate (NMDA) receptor complex. The compounds were evaluated for their affinity at the phencyclidine (PCP) binding site by determining their ability to displace [3H]TCP from crude rat brain synaptic membranes. A wide range of affinities were observed, with the most potent analogs possessing IC50's equivalent to that of the reference agent MK-801 (3, dizocilpine). NMDA antagonist activity was demonstrated by prevention of glutamate-induced accumulation of [45Ca2+] in cultured rat cortical neurons. Selected compounds were also studied in vivo to determine their ability to prevent the lethal effects of systemically injected NMDA in the mouse. In general, the SAR of the phenanthrenamine series may be summarized as follows: (a) for the amino group at C4a, NHMe > NH2 > NHEt >> NC5H10; (b) for the B-ring substitution, X = CH2 > S > O; (c) unsaturation of the C ring decreases receptor affinity; (d) cis-ring fusion between the B and C rings is desirable; (e) 6-hydroxy or 6-methoxy substitution of the phenanthrenamine system identified an additional hydrogen bonding interaction that substantially increased receptor affinity; (f) spiro analogues (such as 55, IC50 = 3400 nM), which altered the point of attachment of the C ring, caused a substantial reduction in PCP-site affinity. Molecules from this series were useful for refining a pharmacophore model consistent with previous models of the PCP site. In this model, the (R)-(+)-phenanthrenamine 13 superimposes closely onto MK-801 (3), and the angular 4a-amino group is believed to hydrogen bond with a putative receptor site atom. In the phenanthrenamine and thiaphenanthrenamine series, the (R)-(+)-enantiomers (9, 13, and 44) are more potent by approximately 5-10-fold than their corresponding (S)-(-)-enantiomers with respect to their affinity for the PCP site, their ability to prevent accumulation of [45Ca2+] in cultured neuronal cells, and their protection against the lethal effects of NMDA in mice. In general, there was no separation between the dose that prevented NMDA lethality and the dose that produced ataxia in mice, except in the case of the thiaphenanthrenamines 41 and 43. We have not yet obtained evidence that this small separation in activity offers a therapeutic advantage in the treatment of cerebral ischemia or other neurodegenerative disorders.