The in vivo efficacies of ceftazidime, aztreonam, and the combinations of ceftazidime with amikacin and aztreonam with amikacin were studied in the rabbit left-sided endocarditis model by using two strains of Pseudomonas aeruginosa, one multisusceptible and one multiresistant, in a total of 156 animals. Antibiotics were given intramuscularly for 10 days, as follows: amikacin, 7 mg/kg of body weight every 8 h, and ceftazidime and aztreonam, 50 mg/kg every 8 h. All regimens except amikacin alone significantly reduced the number of CFU per gram of vegetation (P < or = 0.008), but only for the multisusceptible strain for which sterile vegetations were obtained in 20, 25, 21, 75, and 53% of the groups treated with amikacin, ceftazidime, aztreonam, and the combination groups ceftazidime-amikacin and aztreonam-amikacin, respectively (ceftazidime plus amikacin versus controls, P = 0.001). Regarding the decrease in the numbers of colonies in vegetations, (i) all regimens significantly reduced the number of CFU per gram of vegetation (P < 0.001), (ii) results with ceftazidime-amikacin compared with those with monotherapy were significantly different (P < or = 0.007), and (iii) results with aztreonam-amikacin, although better than those with monotherapy, were marginally not statistically significant. At 1 h postdose, mean amikacin, aztreonam, and ceftazidime levels in serum were 35 +/- 19.4, 89.6 +/- 8.16, and 92.61 +/- 11.52 micrograms/ml, respectively. It was concluded that the combination of ceftazidime, and possibly aztreonam, with amikacin given at high doses and short intervals could have a place in the therapy of patients with left-sided endocarditis caused by P. aeruginosa.