Objectives: This manuscript attempts to critically review traditional and currently employed methods of periodontal diagnosis, in the light of current knowledge about individual patients and sites at risk of progressive periodontal attachment loss. Data sources: Articles published over the last decade from international research journals, have demonstrated that existing methods of periodontal disease diagnosis are seriously deficient with respect to accuracy, their ability to predict ongoing or future disease activity and their ability to determine the current activity status of historically diseased sites. Study selection: Longitudinal studies have questioned the rationale behind traditional treatment regimes and underlined the importance of site-directed therapy to avoid potentially damaging instrumentation of quiescent or healthy sites. The recent explosion in local, less invasive chemotherapies for periodontal disease management has aimed at addressing the site-specific nature of this group of diseases, but the true benefits of such novel therapies cannot be realised until more accurate and specific diagnostic techniques become available. Conclusions: The manuscript concludes that the range of clinical information collected by experienced periodontists using currently available technology is probably sufficient to manage mild-to-moderate chronic adult periodontitis. However, those patients at risk from more aggressive attachment loss, and those individuals that appear refractory to traditional therapies, require the development of more accurate diagnostic tests to compliment the revolution in site-specific therapies. A diagnostic model is presented, which attempts to draw together current and future diagnostic methods for managing the majority of periodontal disease types, and it is suggested that current diagnoses should include some assessment of 'risk'. (C) 1997 Elsevier Science Ltd.