The relation between relative weight and health differs between young and old. In older populations, weight change may cloud the association between a single relative weight and health outcomes. To determine whether weight or weight change is a more important determinant of mortality in a population of older adults, the authors analyzed data from the Systolic Hypertension in the Elderly Program (1984-1990), a randomized clinical trial testing the efficacy of antihypertensive drug treatment to reduce the risk of stroke in older adults (aged 60 years or more) with isolated systolic hypertension. After adjustment for covariates, an average annualized weight loss of at least 1.6 kg/year (odds ratio = 4.9), a weight loss between 1.6 and 0.7 kg/year (odds ratio = 1.7), a weight gain of more than 0.5 kg/year (odds ratio = 2.4), and a baseline body mass index of less than 23.6 (odds ratio = 1.4) all had a significant (p < 0.05) association with all-cause mortality compared with a referent group that was weight stable and of intermediate body mass index (23.6 to <28.0 kg/m(2)) and weight change (-0.7 to <0.5 kg/year). The authors conclude that, in older adults, dynamic measures (e.g., annualized weight change) of weight change predict mortality better than do static weight measures (e.g., baseline body mass index). Even in those with high or low baseline body mass index, weight stability is associated with a lower mortality risk.