Recent experimental work has shown that a so-called PAE (postantibiotic effect, i.e. persistent suppression of regrowth after short exposure of bacteria to the study drug in vitro ) is a feature of most current antibiotics. However, marked quantitative differences were found between different types of antibiotics and also between Gram-positive and Gram-negative organisms studied. A PAE has not yet been demonstrated for roxithromycin, a new macrolide antibiotic. Therefore, we compared the PAE of roxithromycin, erythromycin, and clindamycin against laboratory strains and clinical isolates of Staphylococcus aureus, Streptococcus pyogenes, Str.pneumoniae , and Haemophilus influenzae in vitro . Identical multiples of the MIC and identical exposure times resulted in similar PAEs for the three study drugs tested. Good correlations could be found between the area under the in-vitro concentration-vs-time curve (AUC) and PAEs. The longest PAE of 9.6 h was observed after exposure of Str. pneumoniae to 1.9 mg/l of roxithromycin for 6 h.