The predictive utility of 5 major putative periodontopathic microbial species, ''superinfecting'' organisms, and several clinical periodontal parameters were assessed relative to periodontitis recurrence over a 12-month period in 78 treated adult patients participating in a 3-month maintenance care program. At baseline, pooled subgingival microbial samples were collected from each patient, and whole-mouth evaluations of probing depth, relative periodontal attachment level, furcation involvement, and indices of plaque and gingival inflammation mere carried out. 67 (85.9%) subjects were culture-positive at baseline for presence of either Actinobacillus actinomzycetemcomitans, Porphyromonas gingivalis, Prevotella intermedia, Campylobacter rectus or Peptostreptococcus micros, with 48 (61.5%) subjects yielding one or more of these species at or above designated threshold proportions of greater than or equal to 0.01% for A. actinomycetemcomitans, greater than or equal to 0.1% for P. gingivalis, greater than or equal to 2.5% for P. intermedia, greater than or equal to 2.0% for C. rectus, and greater than or equal to 3.0% for P. micros. Subgingival yeasts were recovered from 12 subjects, staphylococci from 7, and enteric rods/pseudomonads from 6; however, no subjects revealed greater than or equal to 1.0% baseline proportions of these ''superinfecting'' organisms in subgingival specimens. Periodontitis recurrence in subjects was defined as any periodontal site exhibiting either a probing depth increase of greater than or equal to 3 mm from baseline, or a probing depth increase of greater than or equal to 2 mm from baseline together with a loss in relative periodontal attachment of greater than or equal to 2 mm from baseline. 15 (19.2%) study subjects showed periodontitis recurrence within 6 months of baseline, and 25 (32.1%) within 12 months. The mere baseline presence of the 5 major test species and ''superinfecting'' organisms were not significant predictors of periodontitis recurrence over 12 months. However, a 2.5 relative risk for periodontitis recurrence over 12 months was found for subjects yielding one or more of the 5 major test species at or above the designated baseline threshold proportions (p = 0.022, Mantel-Haenszel chi(2) test). The positive predictive value for periodontitis recurrence of a microbiologic analysis encompassing the 5 major test species at or above the designated threshold proportions improved with increasing time from baseline, up to approximately 42% at 12 months. Baseline variables jointly providing in multiple regression analysis the best predictive capability for periodontitis recurrence in subjects over a 12-month period were recovery of one or more of the 5 major test species at or above designated threshold proportions, the proportion of sites per subject with greater than or equal to 5 mm probing depth, and the mean whole-mouth probing depth. These findings indicate that one or more of 5 major putative periodontal pathogens in elevated subgingival proportions together with increased probing depth predispose adults on maintenance care to recurrent periodontitis.