The postantibiotic effect (PAE) of meropenem was determined for 11 strains, both clinical isolates and reference strains of members of the family Enterobacteriaceae. The study compares PAE results obtained by five methods used to monitor bacterial regrowth, including viable counting, alone and in combination with impedance; bioluminescence, alone and in combination with impedance; and a morphological technique. After exposure of the test organisms to meropenem (0.1 x to 100 x MIC) for 2 h, concentration-dependent differences in counts by bioluminescence and viable counts were observed, the latter always being lower. The differences varied with the test organism. For example, after exposure ofProvidentia stuartii NCTC 10318 to 0.1 x MIC, the counts were 5.5 x 105 and 2.0 x l05 whereas after exposure of Citrobacter freundii MR76 to 0.1x MIC of meropenem the counts were 2.3 x 106 and 6.8 x 103 by bioluminescence and viable counting, respectively. The discrepancies were probably due to the relative inability of the viable counting procedure to detect fragile aberrant morphologies and resulted in differences in the calculated PAE values. With methods which do not detect fragile morphologies, the PAE may be underestimated. A general trend was observed for the order of magnitude of the PAEs by the following methods (in order of decreasing magnitude of PAE): (i) morphological technique, (ii) bioluminescence technique alone, (iii) bioluminescence in combination with impedance, (iv) viable counting in combination with impedance, and (v) viable counting alone. It is our opinion that of the methods examined in this study, bioluminescence in combination with impedance best reflects the true values for PAEs, and these results were examined more closely.Postantibiotic effect (PAE), the delayed regrowth of bacteria after antibiotic exposure, is now a well-established phenomenon that was first described almost 50 years ago (1). The effect appears to be a feature of virtually all antimicrobial agents and has been observed for a wide variety of bacteria and yeasts. It is, however, the specific antimicrobial agent-organism combination which determines the presence and duration of the PAE. Investigators using both old and new ,-lactams have consistently observed a PAE with gram-positive cocci (2). In contrast, after exposure of gram-negative bacilli to ,-lactams, PAEs, if any, are of short duration or have a negative value. PAE values obtained are, however, very dependent on the methods used to quantify the effect (8, 12). Bearing in mind that gram-negative bacilli generally form fragile spheroplasts when exposed to meropenem, the suitability of performing viable counts on these cell forms has been questioned, and it was found that an underestimation of the number of cells by viable counting led to an underestimation of the PAE (9,13,15). This study extends these observations of some members of the family Enterobacteriaceae by using an impedance technique (3) in conjunction with a bioluminescence assay of bacterial ATP (6) and compa...