The disposition kinetics and bioavailability of streptomycin, kanamycin and neomycin were determined following their administration as parenteral preparations to horses. Single doses (10 mg/kg) of each aminoglycoside were given by the intravenous (i.v.) and intramuscular (i.m.) routes and, at a later time, seven intramuscular doses were injected at 12-h intervals. The pharmacokinetic behaviour of the three aminoglycosides was similar, in that a rapid distribution phase was followed by a relatively short half-life. The half-life (mean +/- SD, n = 6) of kanamycin (1.80 +/- 0.17 h) was significantly (P less than 0.01; t test, 10 d.f.) shorter than that of streptomycin (3.40 +/- 0.42 h), while neomycin half-life (2.10 +/- 0.97 h) was of an intermediate length. The apparent volume of distribution of neither kanamycin nor neomycin varied significantly (P greater than 0.05) from that of streptomycin and numerically (V1 d congruent to 230 ml/kg) was the same as the extracellular fluid volume. The body clearance of kanamycin (88.5 +/- 11.3 ml/kg.h) was significantly (P less than 0.01) larger than that of streptomycin (47.5 +/- 7.9 ml/kg.h), while a significant difference in this parameter did not exist (P greater than 0.05) between neomycin and streptomycin. Following intramuscular injection, each aminoglycoside was rapidly and completely absorbed from the injection site, although neomycin showed wide individual variation in the fraction absorbed. The administration of multiple doses did not change either the bioavailability or the apparent half-life from the values obtained after a single dose. The only pharmacokinetic difference between these aminoglycosides that is of clinical importance lies in the rate of their elimination. A dosage interval of 8 h would be appropriate for kanamycin compared with a 12-h interval for streptomycin. The dosage interval for neomycin based on half-life should be 8 h but, due to the relatively greater toxicity of this aminoglycoside, an interval of 12 h might be recommended. The height of the peak serum concentration is determined by the size of the dose.