Although many epidemiological studies have been conducted concerning periodontal disease, the majority were not included in this review because of deficiencies in the measures used. Although it is increasingly common for studies in this field to measure periodontal disease using clinical attachment level, attachment loss or bone loss, the evidence pertaining to prevalence, incidence and risk in older adult populations is limited. Although it is the best indicator to date, characterizing periodontal disease by means of attachment loss has some limitations. Prevalence and incidence rates may vary according to the number of teeth and sites probed and bias and case misclassification may occur because of the healthy survivor effect. Moreover, prevalence data that document lifetime disease experience are of little use in planning for periodontal treatment needs. Problems with sampling or subject selection and idiosyncratic ways of reporting data also limit the quality of the evidence currently available. In order to standardize the collection of data on loss of attachment and to measure it as accurately as possible, Papapanou (63) recommends that studies use full-mouth periodontal examinations and the assessment of clinical attachment level at four sites on each remaining tooth. Given the inconsistencies in and problems with the methods used in the studies reviewed above, only broad conclusions can be drawn concerning periodontal disease in older adults. These confirm the conclusions reached in other reviews of the literature. While moderate levels of attachment loss are to be found in a high percentage of middle-aged and elderly subjects, severe loss is confined to a minority, albeit a substantial one. Severe loss is evident in only a few sites and, in general, affects only a small proportion of sites examined. Nevertheless, approximately one-fifth of older individuals have experienced more generalized severe loss; the rate is much higher in the oldest subjects and subjects from minority groups. Although not universal, severe disease is common in some older populations and some population subgroups. Studies using common approaches are needed to fully elucidate the extent to which disease experience varies across different populations. Similar conclusions can be drawn from prevalence studies measuring bone loss. These show that a minority of subjects accounted for most sites with advanced loss. Studies of incidence suggest that 50-75% of older adults experience additional loss of attachment of 2 or 3 mm or more at a minimum of one site over relatively short periods of time. Rates fall dramatically when more stringent case definitions are used. Moreover, relatively few sites examined show evidence of additional loss so that, although rates are high, extent and severity are low. More detailed analyses of incidence data, although few, indicate that new lesions are more common than progressing lesions, and the pattern of loss tends to support an episodic model of periodontal disease progression.