Since in-vitro methods for studying drug interactions are difficult to evaluate and different results occur with the chequerboard (isobolograms-FIC index) and killing curve methods, a new approach associating these two methods was used to study the ceftazidime-pefloxacin interaction. In 64 tubes in a chequerboard pattern, viable bacteria were counted at 0, 2.5, 5 and 24 h, by a microdilution method and a multiple inoculator replicating method. The bacterial inoculum consisted of 5 X 10(6)-10(7) cfu/ml. Twelve strains belonging to six genera were tested: Acinetobacter, Citrobacter, Enterobacter, Klebsiella, Serratia and Pseudomonas. During the first 5 h no antagonism was noted, but few differences in viable count equal to or greater than 2 log10 were observed. The results of the drug interaction were equivalent to those of the more rapid bactericidal antibiotic concentrations (ceftazidime or pefloxacin). At 24 h, synergy was observed: ceftazidime prevented the late regrowth noted with pefloxacin, but with Citrobacter spp. regrowth was also observed with ceftazidime and none of the combinations were bactericidal.