Objective: To explore the association between global self-rated health and mortality in older people. Methods: A systematic review was performed. The inclusion criteria were longitudinal studies that assessed self-rated health with a single general question and samples of community-dwelling persons aged 60 years or more. Electronic databases were searched and references were reviewed. Results: We selected 18 studies published between 1993 and 2011. Six out of seven studies that analyzed men and women found a higher risk of dying among persons who rated their health as poor the most frequent covariables were age, gender, chronic diseases, and functional status. Half of the studies that analyzed only men or women found a significant association. The effect of self-reported health on mortality was observed among people younger than 75 years. Results were not dependent on the length of follow-up. Conclusions: The results confirm previous findings suggesting that a negative self-rating of g