Iminodibenzyl-, iminostilbene-, dibenzocycloheptadiene-, dibenzooxepine- and dibenzothiepine-derivatives of tricyclic antidepressant drugs were able to inhibit Na+-stimulated Mg2+ efflux in human erythrocytes at concentrations of 10(-5) -10(-3) mol/l. Tricyclic antidepressant drugs belonging to other chemical groups, non-tricyclic antidepressant drugs and phenothiazines were less potent inhibitors (IC50 of 10(-4) mol/l or higher). Imipramine and dothiepine, the most potent compounds, inhibited the Mg2+ carrier with IC50 of 2.5 and 4 x 10(-5) mol/l respectively. These IC50 are of similar order of magnitude to those of some classical transport inhibitors (such as furosemide for the [Na+ K+, Cl-]-cotransport system). In addition, these concentrations of imipramine and dothiepine were free of: i) side effects on other erythrocyte Na+ and K+ transport pathways (with the exception of a slight inhibition of Ca2+-sensitive K+-channels and [Na+,K+,Cl-]- and [K+,Cl-]-cotransport systems) and ii) toxic effects on the membrane leak for divalent or monovalent cations. Therefore, we selected imipramine as an useful tool for investigating fluxes catalyzed by the Na+-stimulated Mg2+ carrier. Imipramine was tested on the initial rate of ouabain and bumetanide-resistant net Na+ influx in Na+-depleted, Mg2+-loaded erythrocytes. The compound was able to inhibit a Na+ influx of about 300-500 mumol (1.cells x h)-1 with an IC50 of about 3 x 10(-5) mol/l. This imipramine-sensitive Na+ influx was coupled with an imipramine-sensitive Mg2+ efflux in a stoichiometry of 3.03 +/- 0.34 (mean +/- SEM of 7 experiments).