OBJECTIVES: To determine whether elder abuse can predict mortality and disability over the ensuing 12 years. DESIGN: Population-based prospective cohort study of women aged 70 to 75 in 1996; survival analysis. SETTING: Australia. PARTICIPANTS: Twelve thousand sixty-six women with complete data on elder abuse. MEASUREMENTS: Elder abuse was assessed using the 12-item Vulnerability to Abuse Screening Scale (VASS) subscales: vulnerability, coercion, dependence, and dejection. Outcomes were death and disability (defined as an affirmative response to "Do you regularly need help with daily tasks because of long-term illness, disability or frailty?"). RESULTS: In 1996, 8% reported vulnerability, 6% coercion, 18% dependence, and 22% dejection. By October 2008, 3,488 (29%) had died. Mortality was associated with coercion (hazard ratio (HR) = 1.21, 95% confidence interval (CI) = 1.06-1.40) and dejection (HR = 1.12, 95% CI = 1.03-1.23), after controlling for demographic characteristics, social support, and health behavior but not after adding chronic conditions to the coercion model. Over the 12 years, 2,158 of 11,027 women who had reported no disability in 1996 reported disability. Women who reported vulnerability (HR = 1.25, 95% CI = 1.06-1.49) or dejection (HR = 1.55, 95% CI = 1.38-1.73) were at greater risk of disability, after controlling for demographic characteristics, social support, and health behavior. The relationship remained significant for dejection when chronic conditions and mental health were included in the model (HR = 1.40, 95% CI = 1.24-1.58). CONCLUSION: Specific components of vulnerability to elder abuse were differently associated with rates of disability and mortality over the ensuing 12 years.