The aim of the present study was: (1) to assess longitudinally the occurrence of Actinobacillus actinomycetemcomitans (Aa) in young subjects wearing fixed orthodontic appliances compared to matched appliance-free controls; (2) to determine whether the presence of the micro-organism at baseline could influence the periodontal status assessed 3 years later. 70 subjects, 27 male and 43 female, aged between 12 and 20 years participated in the study: 35 subjects under orthodontic treatment with fixed appliances for at least 6 months, and 35 appliance-free individuals matched for age and gender. All subjects were free of clinically demonstrable loss of attachment. They all received oral hygiene instructions 2x during the 2 months preceding the first clinical and microbiological examination. No subgingival instrumentation was performed between baseline and the 3-year examination. Clinical parameters included gingival bleeding index (GBI), pocket probing depth (PPD) and measurements of attachment level (AL). Statistically significant differences were reported regarding frequency of detection of Aa between both groups at each examination. The %s of orthodontic subjects infected with Aa at the baseline and at the 3-year examination were 86% and 80%, respectively, while the corresponding figures for control subjects were 16.6% and 26.6%. The frequency distribution of %s of Aa in the total anaerobic subgingival flora among control subjects remained fairly stable, whereas the proportion of orthodontic subjects yielding Aa at a concentration > or = 1.0% dropped significantly from 32% at baseline to 19% at the 3-year visit. Calculations of the relative risk for increasing GBI and PPD in both groups when Aa was present at baseline, revealed that the orthodontic subjects positive for Aa had a negligible relative risk of experiencing worse periodontal conditions compared to orthodontic patients where Aa was not detected at baseline. In contrast, control subjects initially infected with Aa presented with a risk for increased GBI 6.6x higher than that for subjects without Aa. In conclusion, the present study confirmed previous cross-sectional findings reporting that young individual with an integer periodontium wearing fixed orthodontic appliances harbor Aa with a statistically significant greater frequency than appliance-free matched controls. However, although orthodontic patients exhibited more inflammation, their deteriorated clinical conditions could not be accounted for by the sole presence of Aa in their sulci. In contrast, appliance-free young subjects initially infected with Aa had a higher risk of experiencing more gingival inflammation than subjects without the bacterium during a 3-year observation period.