The effect of various volumes of blood and serum on the in vitro antibacterial activity of 0.2% nitrofurazone soluble dressing (NSD) was studied. The antibacterial activity of NSD was tested with an agar well diffusion technique. Zones of inhibition against susceptible strains of Escherichia coli and coagulase-positive Staphylococcus aureus were measured with a micrometer. Bacterial concentrations of 10(5) and 10(8) colony-forming units per milliliter were tested to evaluate a possible effect of inoculum size. Wells contained full-strength (undiluted) NSD or 75%, 50%, or 25% NSD dilutions (w/w) in blood, serum, or 0.9% sodium chloride injection. The mean decrease in inhibition zone size produced by blood and serum was only 7.2%. The diminution of activity, albeit small, was statistically significant. The zones produced were still much larger than those associated with clinical cure. Therefore, the impact of blood and serum on nitrofurazone's in vitro antibacterial activity can best be described as a slight reduction in rather than an elimination of effectiveness. This small reduction in activity is unlikely to be clinically important in patients with burns or other surface wounds that contain blood or serum. An effect of inoculum size was demonstrated for both organisms. Blood and serum produced a small but significant reduction in NSD's antibacterial activity in vitro. Controlled clinical studies are needed to ascertain the clinical importance of these findings.