A field study using four different private periodontal practices and two independent laboratories was conducted to compare two microbiologic cultures sampled simultaneously from the same sites in 20 individual patients. Both paired samples were submitted separately to one of the two independent laboratories for bacterial identification and antibiotic sensitivity testing. The results from the two samples were quite variable. In only two instances did both specimens reveal the presence of identical bacterial species, but these specimens differed in both threshold levels and antibiotic sensitivity. When only bacteria above threshold levels were compared, total agreement was found in 11 of 20 cases. When examining antibiotic sensitivity, using 100% kill as the ideal, agreement between the two specimens was inconsistent. The use or nonuse of tetracycline was in agreement 85% of the time, amoxicillin 75% of the time, metronidazole 70% of the time, and amoxicillin-metronidazole in combination 85% of the time. The two specimens agreed on the empirical use of amoxicillin 45% of the time, tetracycline 60% of the time, and metronidazole 60% of the time. The empirical use of amoxicillin-metronidazole in combination yielded 80% agreement when the results of both specimens were combined. The empirical use of amoxicillin-metronidazole combination therapy may be more clinically sound and cost effective than culturing and antibiotic selection based on the culture from any single microbiologic testing laboratory. This supports the data from a previous study that examined specimens secured simultaneously from the same site and submitted to two different testing laboratories. The failure of microbial testing to achieve a higher level of consistency between samples leaves the clinical efficacy of microbial testing in question.