We prospectively evaluated serum aminoglycoside (AMG) concentrations in 120 patients who received gentamicin or tobramycin for serious gram-negative bacillary disease. AMG serum concentrations were assayed by microbiological and radioimmunoassay techniques. Correlation between the two assay methods was good. When AMG doses were based on total body weight, there was no significant correlation between AMG dosage administered and serum concentrations in patients with either normal or abnormal renal function. The use of ideal body weight for calculation of AMG dosage improved this correlation significantly except in hemodialysis patients. AMG-induced nephrotoxicity occurred in 13 patients. No significant association was noted between the occurrence of toxicity and the specific AMG given or with other commonly recognized risk factors. Among study groups, peak AMG serum concentrations failed to exceed the minimal inhibitory concentration of the infecting organism in 17 to 33% of the cases. Serum inhibitory levels of ≥ 1:8 were not associated with improved survival. There was no significant difference in mortality between the gentamicin- and tobramycin-treated groups. We advise baseline serum AMG levels in seriously ill patients with gram-negative bacillary disease and additional bacteriological studies in selected situations.