The influence of technical factor variations on serum bactericidal and serum inhibitory titers was studied by using Staphylococcus aureus clinical isolates versus oxacillin-spiked human serum. Parallel tests, both with and without the use of beta-lactamase in count plates to inactivate oxacillin carryover, were performed with a conventional macrodilution approach, a carefully controlled macrodilution procedure, and a standard microdilution method. Careful control of technical factor variations diminished the incidence of low serum bactericidal titers and decreased the dispersion of results, a finding corollary to the known influence of technical factor variations on the measurement of MBCs. The incorporation of beta-lactamase into count plates resulted in a shift of serum bactericidal titers to lower values. The microdilution method appeared to be least influenced by technical variations and, with the addition of beta-lactamase to count plates, provided the best results.