This study describes some predictors of new and progressing destructive periodontal disease over a 10-year period in rural Chinese. A total of 398 persons aged 20-80 years, who had participated in a baseline survey of tooth mortality, dental caries and periodontal conditions and were still dentate 10 years later, were reexamined for the same parameters as assessed at baseline. Three different threshold values were used to define new and progressing destructive periodontal disease. Irrespective of the threshold used, most persons experienced new disease. Progressing disease was very prevalent when a 2+ mm disease definition was used, but occurred less frequently at the higher threshold levels. The logistic regression models for 2+ and 3+ mm disease were essentially similar, and showed that women, persons with 104 sites or more, and persons with 0-5% sites with 4+ mm attachment levels had a lower risk of disease progression as well as of new disease than did men, persons with few sites and persons with 6% sites or more with attachment levels 4+ mm. The variables sex, number of sites present, percentage of sites with 4+ mm attachment levels, and presence of mobile teeth were predictive for new disease using a 4+ mm definition. Age, percentage of sites with 4+ mm attachment levels and percentage of sites with 4+ mm pockets were predictive for progressing disease using the 4+ mm disease definition.