PURPOSE Periodontitis has been linked to the occurrence of various systemic diseases. The goal of this study was to explore the periodontitis-cancer association in the NHANES I Epidemiologic Follow-up Study. METHODS Data were available on 11,328 adults, age 25 to 74 years, who were diagnosed as dentate individuals with either periodontitis (n = 2092), gingivitis (n = 2603), a healthy periodontium (n = 2,671), or as individuals without teeth (edentulous n = 3,962) at the beginning of the follow-up. The main outcome measure was fatal cancer, as ascertained from death certificates. RESULTS Compared with individuals with a healthy periodontium, fatal cancer occurrence was positively associated with periodontitis at baseline (age and gender adjusted odds ratio = 1.55, 95% confidence interval: 1.25-1.92). Of the different cancer types, lung cancer demonstrated the strongest association. After adjustment for known risk factors for lung cancer, the magnitude of the association between periodontitis and lung cancer ranged between 1.48 (95% confidence interval: 0.88-2.50) and 1.73 (95% confidence interval: 1.01-2.97). CONCLUSIONS Associations between periodontitis and lung cancer mortality can be identified above and beyond adjustment for known risk factors for lung cancer. Despite these apparent unconfounded associations, there are reasons to believe that the periodontitis-cancer associations may be spurious.