Individual randomized trials comparing aminoglycosides have usually involved a sample size inadequate for reliable detection of small to moderate differences. In a search for overall effects that might not be apparent from any single trial, a quantitative overview of all randomized studies comparing amikacin, gentamicin, netilmicin, sisomicin, and tobramycin was performed. These aminoglycosides appeared equally efficacious, except that tobramycin seemed less effective than sisomicin. Patients given gentamicin or sisomicin seemed to run a higher risk of nephrotoxicity than those given amikacin; the risk for patients receiving tobramycin was lower than that for patients given gentamicin but higher than that for those given netilmicin. The only statistically significant difference with regard to auditory toxicity was a lower risk among patients receiving netilmicin than among those given amikacin or tobramycin. Meaningful differences in toxicity may exist for individual aminoglycosides given in efficacious doses. Future trials must include sample sizes sufficient to detect differences of the magnitudes observed in this overview. In the absence of such data, the overview may guide clinicians in determining the risk-to-benefit ratio for a particular aminoglycoside, especially for patients at high risk of toxicity.