A considerable amount of circumstantial evidence indicates that most forms of periodontitis are due to the presence or dominance of a finite number of bacterial species in the subgingival plaque. Almost all of the putative pathogens are anaerobic species, indicating that most forms of periodontitis could be diagnosed as anaerobic infections. In this double-blind investigation, patients with elevated proportions or levels of spirochetes in 2 or more plaque samples, i.e., 60% spirochetes, were randomly assigned to receive either metronidazole, 250 mg 3 x a day for 1 week, or placebo (positive-control) after the completion of all debridement procedures. When the patients were re-examined 4 to 6 weeks later, the patients in the metronidazole group (n = 15) exhibited a highly significant (p less than 0.01) reduction in probing depth and apparent gain in attachment levels relative to the patients (n = 18) in the positive-control group about those teeth that initially had probing depths of 4 to 6 mm. This pattern was also observed about teeth that initially had probing depths greater than or equal to 7 mm. This reduction in probing depths and apparent gain in attachment was associated with a significant reduction in the need for periodontal surgery in the metronidazole-treated patients (difference 8.4 teeth per patient) compared to the positive-control patients (2.6 teeth per patient). These clinical improvements in the metronidazole group were associated with significantly lower proportions of spirochetes, selenomonads, motile rods, and P. intermedius, and a significantly higher proportion of cocci in the plaques. These findings indicate that systemic metronidazole, when given after all the root surface debridement is completed, leads to additional treatment benefits, including a reduced need for surgery, beyond that which can be achieved by debridement alone.