Disease progression, as manifested by loss of attachment, may be due to an inability of the host to eliminate infecting pathogens from subgingival plaque. Hence, antibody responsiveness to infecting pathogens could represent a useful diagnostic indicator of future attachment loss. Objective: This study looked at the usefulness of an infection ratio (serum antibody response to concurrent bacterial colonisation) as an indicator of attachment loss over the following year. Methods: Subgingival plaque (12 sites/person) and blood samples were collected from 414 individuals at yearly intervals for 5 years. An ELISA was used to analyse plaque samples for P. gingivalis (P.g), A. actinomycetemcomitans (A.a) and P. intermedia (P.i) and to determine the serum IgG, IgA and IgM response to the same species. Infection ratios (IR) were computed for each subject (antibody titre/bacterial load) and compared in those who did or did not lose attachment (≥2mm in ≥2 sites) over the following year. Results: Infection ratios were generally low (median values 0-0.75) for each bacterial species across the 5 years in both groups of subjects. For P.g, median IR values for IgM were higher (0.6-0.75) than for IgA and IgG (0-0.65), with IgG showing more variability. A similar pattern was observed for A.a while median IR values for P.i IgG and IgM were more uniform. However, there was no difference in IRs for any bacterial species in those who did or did not lose attachment. Conclusion: Infection ratios were not found to be a useful indicator of attachment loss over the next twelve months in this population. This study was supported by Colgate Oral Care Australia and Colgate Palmolive Company USA.