Background: The validity of the surveys on self-reported smoking status is often questioned because smokers underestimate cigarette use and deny the habit.It has been suggested that self-report should be accompanied by cotinine test.This report evaluates the usefulness of serum cotinine test to assess the association between smoking and periodontal status in a study with a large sample population to be used in studies with other serum markers in epidemiologic and periodontal medicine researches.Material and Methods: 578 patients who were part of a multicenter study on blood biomarkers were evaluated about smoking and its relation to periodontal disease.Severity of periodontal disease was determinate using clinical attachment loss (CAL).Smoking was assessed by a questionnaire and a blood sample drawn for serum cotinine determination.Results: The optimal cut-off point for serum cotinine was 10 ng/ml.Serum cotinine showed greater association with severity of CAL than self-report for mild-moderate CAL [OR 2.03 (CI95% 1.16-3.53)vs. OR 1.08 (CI95% 0.62-1.87)] advanced periodontitis ] and extension of CAL > 3 mm ].When the two tests were evaluated together were not shown to be better than serum cotinine test.Conclusions: Self-reported smoking and serum cotinine test ≥ 10ng/ml are accurate ,complementary and more reliable methods to assess the patient's smoking status and could be used in studies evaluating serum samples in large population and multicenter studies.Clinical Relevance: The serum cotinine level is more reliable to make associations with the patient's periodontal status than self-report questionnaire and could be used in multicenter and periodontal medicine studies.