To participate effectively in their health care, older patients often are expected to read a wide variety of materials, including written instructions, brochures, and consent forms. This study quantitates the reading ability of older patients and compares it to that of younger patients. Two hundred seventy-two patients 30 and older were selected from five outpatient clinics at a public teaching hospital and tested for objective reading ability using the Peabody Individual Achievement Test--Revised. The 76 patients 60 and older read significantly worse (grade level 2.9) than the 196 patients younger than 60 (grade level 5.8) (P < 0.0001). Older patients also completed significantly fewer years of school than younger patients (7.3 years versus 10.6 years). Analysis of variance for age categories 30-44, 45-59, 60-74, and 75 and older confirmed declining reading ability and educational status with advancing age. Multiple regression analysis helped show that an equation could be derived to predict reading ability from age, educational status, race, and sex, but the coefficient of determination was so low (r2 = 0.39) that it cannot be considered clinically useful for individual patients. In this study, older patients read significantly worse than younger patients, and a formula that combines age, race, sex, and educational status cannot reliably predict reading ability for individual patients. Most older patients read on a level so low that they cannot be expected to read most commonly used written materials. Routine testing of reading ability may allow more appropriate design and use of written materials.