The relationship between probing attachment changes in treated periodontal pockets and the prevalence of selected periodontal pathogens was assessed in 10 patients with adult periodontitis 1 year following randomized therapy. All patients had at least 1 tooth in each quadrant with an inflamed pocket of probing depth > or =5 mm and clinical attachment loss and harbored at least one of the following 3 major periodontal pathogens: Actinobacillus actinomycetemcomitans, Porphyromonas gingivalis, or Bacteroides forsythus. The number of target organisms per site was determined preoperatively; at 1 week; and at 1, 3, 6, and 12 months postoperatively utilizing DNA probes. The following clinical parameters were measured and recorded preoperatively and at 1, 3, 6, and 12 months post-treatment: gingival fluid flow, gingival index, plaque index, probing depth, probing attachment level, gingival recession, and bleeding on probing. One quadrant in each patient was randomly assigned to 1 of the following 4 treatments: 1) scaling and root planing; 2) pocket reduction through osseous surgery and apically-positioned flap; 3) modified Widman flap; and 4) modified Widman flap and topical application of saturated citric acid at pH 1 for 3 minutes. All 4 treatments were rendered in one appointment using local anesthesia. No postoperative antibiotics were used, but patients rinsed with 0.12% chlorhexidine for the first 3 months postoperatively and received a prophylaxis every 3 months. This investigation revealed: 1) 30.0% of the sites were infected by at least 1 species at 3, 6, and 12 months postoperatively. 2) Failing sites were infected by a high number of both Pg and Bf These sites had a mean of 24.2+/-9.0 x 10(3) Pg and 93.1+/-42.0 X 10(3) Bf while stable sites had a mean of 6.8+/-0.5 x 10(3) Pg and 7.2+/-1.2 x 10(3) Bf (P = 0.06 and P = 0.05, respectively). 3) The infected sites lost significantly more mean clinical attachment at 12 months (1.5+/-0.5 mm compared to a loss of 0.2+/-0.3 mm for uninfected sites, P = 0.017). 4) The infected sites had a significantly greater BOP (67+/-14% versus 25+/-8% for uninfected sites at 12 months, P = 0.012). 5) The choice of treatment modality did not affect the prevalence of the target species at 1 year post-treatment. These results suggest that prevalence of microbial pathogens negatively affects the 1 year outcome of periodontal surgical and nonsurgical therapy.