The hydrolysis of BANA by subgingival plaque samples is associated with the presence of either Treponema denticola, Porphyromonas gingivalis, and/or Bacteroides forsythus. A protocol in which pure cultures were incubated for 15 min at 55 degrees C detected about 5 x 10(5) CFU of P. gingivalis and 1 x 10(6) CFU of T. denticola. Clinical studies indicated that the BANA test in this configuration will detect about 10(4) organisms in vivo as compared with the 10(5) to 10(6) organisms found with in vitro grown cells. The BANA test can be made less sensitive by decreasing the time and/or temperature of incubation, which could improve the specificity of the test. In the present study we determined the incubation parameters that would give optimal specificity when the plaque samples were removed from sites of gingival health. Twenty-six approximal plaque samples were taken from each of 90 clinically healthy subjects and incubated with the BANA substrate on PerioScan cards (Oral-B Laboratories) for 5 and 15 min at 35 degrees, 45 degrees, and 55 degrees C. Subjects were randomly assigned to the various temperatures. Wooden toothpicks were inserted interproximally in all sites anterior to distal of the first molars and then each side of the toothpick was wiped onto the PerioScan card. The specificity of the BANA test relative to clinical health was 96% when the cards were incubated for 5 min at 35 degrees C, but decreased to 50-70% when the cards were incubated for 15 min at 35 degrees C or for 5 and 15 min at 45 degrees C and 55 degrees C.(ABSTRACT TRUNCATED AT 250 WORDS)