In this paper, records of all medical conditions on death certificates are used to evaluate hypertension-related mortality in North Carolina over the decade 1968-1977. Use of both an inclusive hypertension recode category and multiple cause data resulted in gains in information of over 750 per cent in all four race/sex groups compared to the commonly used underlying cause, hypertensive disease category. Race, sex and age specific 10-year trends in death rates for all mentions of hypertension are analyzed, with comparisons to underlying cause mortality from ischemic heart disease and stroke. Age-adjusted declines of 19 to 24 per cent between 1968 and 1977 were observed for all race/sex groups, although non-White declines occurred mainly at younger ages while White declines (especially White males) occurred mainly at older ages. The non-White excess of hypertension mentions (compared to Whites) increased for males and decreased for females. The decline in hypertension mentions, in spite of the increased awareness of hypertension as a public health problem which would make it more likely to be mentioned on death certificates, suggests that there was a real reduction in the contribution of hypertension to total mortality over the period.