Commercially available gargles containing (hexetidin 0.1%, dequaliniumchloride 0.01%, benzydamin 0.15%) or not containing (OdolTM 0.5%) antibacterials as well as hydrogen-peroxide (3% ml/ml) and aqua dest. as a control tested for their antibacterial efficacy in vivo by applying them for 30 s. Standardized mouth- and throat-washings of 15 volunteers were examined for their aerobic viable counts before as well as 5 and 60 min after gargling. The 5-minutes values (tab. 1) indicated only hexetidin to reduce the oro-pharyngeal flora significantly though not very profoundly by 0.95 orders of magnitude. Usage of the other gargles caused only log reductions of 0.16 to 0.39, aqua dest. such of 0.23. Sixty minutes after treatment (tab 2.) there was still a small but significant reduction of 0.61 log units demonstrable with hexetidin and the viable counts after treatment with hydrogen-peroxide were 0.31 log units below the baseline. The log reductions as assessed with the other gargles ranged 0.0 to 0.18. It is concluded that the favourable therapeutic effects observed clinically with many "antiseptic" gargles with infections of the upper respiratory tract are not based on the antibacterial efficacy of such preparations.