The homogeneous substrate-labeled fluorescent immunoassay for tobramycin in serum was evaluated. Standard curves are linear and reproducible. The within-run coefficient of variation for control samples containing low, medium and high concentrations of tobramycin ranged from 1.9 to 6.5%; the between-run coefficient of variation was 11% for a control sample at low concentration. The accuracy measured on spiked samples was good and comparable to the accuracy measured by a microbiological assay; it was somewhat better than for the radioimmunoassay. For patient samples, the results with this method correlated well (r - 0.914) with results obtained by radioimmunoassay.