The pharmacokinetic behaviour of aminosidine (15 mg kg-1) and antimony (25.65 mg kg-1 as N-methylglucamine antimoniate), administered subcutaneously either separately or together was studied on four dogs. The results demonstrated that antimony (Sb) did not significantly modify the kinetics of aminosidine (AM) but that the kinetic behaviour of the metal was markedly influenced by the antibiotic, as shown by the differences in mean residence time (MRT), elimination rate constant (Kel) and area under the curve (AUC) with and without the antibiotic (MRT[Sb] = 243.8 +/- 29.5 minutes, MRT[Sb+AM] = 1067.9 +/- 199.2 minutes; Kel[Sb] = 0.008 +/- 0.001 min-1, Kel[Sb+AM] = 0.0015 +/- 0.0003 min-1; AUC[Sb] = 21,024.6 +/- 4448.5 micrograms min ml-1, AUC[Sb+AM] = 130,478.5 +/- 30,481.7 micrograms min ml-1). The persistence of high serum concentrations of antimony when it was administered with aminosidine suggests that the therapeutic doses commonly used should be reduced and that the interval between administration should be increased to avoid the metal reaching toxic concentrations.