The in vitro activity of difloxacin against canine bacterial isolates from clinical cases was studied in the United States and The Netherlands. Minimal inhibitory concentrations (MIC), the postantibiotic effect, the effect of pH on antimicrobial activity, and the bacterial killing rate tests were determined according to standard techniques. The MICs of American and Dutch isolates agreed in general. The MICs of the American gram-negative isolates ranged from 0.06 to 2.0 microg/ml, and the MICs of the Dutch gram-negative isolates ranged from 0.016 to 8.0 microg/ml. A few European strains of Proteus mirabilis and Klebsiella pneumoniae had relatively high MICs. Bordetella bronchiseptica also was less susceptible to difloxacin. The MICs of the American gram-positive cocci ranged from 0.125 to 4.0 microg/ml, and the MICs of Dutch isolates ranged from 0.125 to 2.0 microg/ ml. Difloxacin induced a concentration-dependent postantibiotic effect that lasted 0.2-3 hours in cultures with Escherichia coli, Staphylococcus intermedius, Streptococcus canis, Proteus spp., and Klebsiella pneumoniae. There was no postantibiotic effect observed against canine Pseudomonas aeruginosa. Decreasing the pH of the medium increased the MIC of Proteus mirabilis for difloxacin. The MICs of Escherichia coli and Klebsiella pneumoniae were lowest at neutral pH and were slightly increased in acid or alkaline media. At a neutral pH, most tested bacterial species were killed at a difloxacin concentration of 4 times the MIC. Similar results were obtained when these same bacteria were tested against enrofloxacin. A Klebsiella pneumoniae strain in an acidic environment was readily killed at difloxacin or enrofloxacin MIC, but at neutral pH the drug concentration had to be raised to 4 times the MIC for a bactericidal effect. After 24 hours of incubation at pH 7.1, difloxacin and enrofloxacin had similar bactericidal activity for all bacteria tested except Staphylococcus intermedius. Against S. intermedius, difloxacin was more bactericidal than enrofloxacin.