A population survey in 1982 confirmed that Nauruan adults suffer from an extremely high prevalence (24%) of non-insulin-dependent diabetes mellitus. A follow-up study of the survey subjects was conducted in 1986. The aim was to assess the burden of diabetes to Nauruans in terms of premature mortality. Age-adjusted mortality rates for diabetic subjects were significantly increased when compared with normal subjects (relative risks for 4-yr mortality were 4.53 in men, P less than .01, and 3.96 in women, P less than .05). Although there was an excess number of deaths among diabetic subjects compared with normal subjects and subjects with impaired glucose tolerance, there was no significant association between cause of death and diabetes.