Chronic (14 day) but not acute (1 day) treatment of mice with clinically active antidepressants produces a significant (approximately 1.8-4.3 fold) reduction in the potency of glycine to inhibit [3H]-5,7-dichlorkynurenic acid (5,7-DCKA) binding to strychnine-insensitive glycine receptors in neocortical membranes. Moreover, these effects were not observed following chronic treatment with a variety of nonantidepressant drugs such as D-deprenyl, chlorpromazine, salbutamol, scopolamine and chlordiazepoxide. The time course and dose-response relationships for this effect were examined after treatment with two representative antidepressant drugs (imipramine and citalopram) and electriconvulsive shock (ECS). Increases in the IC50 of glycine to inhibit [3H]-5,7-DCKA binding were observed after treatment for 7 days with ECS, 10 days with citalopram and 14 days with imipramine, respectively, and were no longer apparent by the 10th day after cessation of treatment. These findings indicate that the antidepressant-induced reduction in the IC50 of glycine to inhibit [3H]-5,7-DCKA binding is 1) a slowly developing, adaptive phenomenon; 2) remarkably persistent after cessation of treatment; and 3) a significantly better predictor of antidepressant activity (22 of 23 drugs) than either beta adrenoceptor down-regulation (15 of 23 drugs) or efficacy in the forced swim test (13 of 23 drugs) [P < .01 vs. each measure, Fisher's Exact Test]. The ability of antidepressants drawn from every principal therapeutic class to effect adaptive changes in the N-methyl-D-aspartate receptor complex is consistent with the hypothesis that this ligand-gated ion channel serves as a final common pathway of antidepressant action and indicates that glutamatergic pathways may be involved in the pathophysiology of depression.