In rats trained to retain a passive avoidance response or to retrieve a learned task in the radial and water maze tests, a pretreatment with 2-hexyl-3-indoleacetamide (FGIN-1-27) (IC50 57 mumol/kg p.o.) or 4' chlorodiazepam (4'CD) (15 mumol/kg i.p.), two steroidogenic ligands at the mitochondria diazepam-binding inhibitor receptor complex (MDRC), antagonized the performance deficit elicited by dizocilpine (0.3 mumol/kg i.p.), a noncompetitive N-methyl-D-aspartate (NMDA) receptor antagonist. The 1-(2-chlorophenyl)-N-methyl-N-(-1-methyl-propyl)-3-isoquinoline carboxamide (PK-11195), an antagonist at MDRC in vivo, failed to modify the disruptive effect of dizocilpine in the passive avoidance response but reversed the FGIN-1-27- or 4' CD-induced antagonism of dizocilpine behavioral actions. Pretreatment with pregnenolone sulfate (48 mumol/kg i.p.), 3 alpha, 21-dihydroxy-5 alpha-pregnan-20-one (THDOC) (15 mumol/kg i.v.) and 3 alpha-hydroxy-5 alpha-pregnan-20-one (allopregnanolone) (15 mumol/kg i.v.) also reduced the passive avoidance retention deficit elicited by dizocilpine. The (17-beta)-17-[[bis(1-methylethyl)amino[carbonyl]androstane-3,5-diene-3- carboxylic acid (SKF-105111), a 5 alpha-reductase inhibitor, blocked the antagonism of dizocilpine behavioral actions by pregnenolone sulfate or by FGIN-1-27 but not those caused by THDOC or allopregnanolone either in normal or adrenalectomized-castrated rats. Thus, it is inferred that the amnesic effect of dizocilpine is counteracted by FGIN-1-27, 4'CD and pregnenolone sulfate because of their ability to increase brain accumulation of allopregnanolone.