Elevated glucocorticoid levels have been associated with cognitive impairment, including dementia. However, few longitudinal studies have examined the association between resting cortisol levels and the incidence of cognitive impairment. We measured overnight urinary excretion of cortisol in 538 high-functioning men and women, 70-79 years of age, in 1988, and assessed their cognitive functioning in 1988, 1991, and 1995 using the short portable mental status questionnaire (SPMSQ). Compared to participants in the bottom quartile of urinary cortisol at baseline, those in the top three quartiles had higher risk of incident cognitive impairment over the 7-year follow up (i.e., decline in SPMSQ score to below six out of nine). This association was not affected by adjustment for age, gender, education level, ethnicity, smoking, prevalent cardiovascular disease, and blood pressure (adjusted odds ratio for the highest quartile 2.34, 95% confidence interval, 1.07-5.14). There was no effect modification by gender; the association was equally strong in men and women. We conclude that urinary excretion of cortisol predicts incident cognitive impairment in older men and women.