Aging is associated with loss of height, which is thought to be caused, at least in part, by vertebral osteoporosis. Estrogen replacement therapy prevents osteoporosis in postmenopausal women, and it has been shown to protect against height loss in some, but not all, prior studies. We took advantage of the availability of information on height measured under standardized conditions 20 years ago among women who participated in the Walnut Creek Contraceptive Drug Study to assess the relationship between height loss and long-term postmenopausal estrogen replacement therapy. Subjects were 36 long-term estrogen users (mean duration of use 17.2 years) and 34 nonusers who were age 45-54 years at entry to the Walnut Creek Study in 1969-1971. These women had their height remeasured in 1992 and completed a detailed questionnaire about risk factors for osteoporosis. After 21-23 years of followup, height loss among estrogen users was 0.77 in (95% C.I. 0.50, 1.04); among nonusers, height loss was 0.94 in (95% C.I. 0.71, 1.17). The difference in mean height loss between estrogen users and nonusers was 0.17 in (95% C.I. - 0.18, 0.52), which was not statistically significant (p = 0.32). After adjustment for factors related to osteoporosis and baseline height, the difference in height loss between estrogen users and nonusers was still 0.17 in (95% C.I. - 0.29, 0.62), a difference that was still not statistically significant (p = 0.46). Height loss was highly variable in both estrogen users and nonusers. Median height change and the distribution of height change were essentially identical between estrogen users and nonusers. Our findings do not contradict the protective effect of postmenopausal estrogen replacement therapy against spinal bone loss. They suggest that estrogen may not prevent biologically important differences in height loss up to age 70.