Perhaps because negative emotions are frequently expressed in physiological reactions, psychosomatic theories have often identified Neuroticism and its component traits (including anxiety, anger, and depression) as causal influences on the development of disease. These views are apparently supported by correlations between physical symptom reports and measures of Neuroticism in males. Data from 347 adult women in the Baltimore Longitudinal Study of Aging replicate this finding for total physical complaints and for most body systems. However, analyses of mortality in the literature and in the present article show no influence of Neuroticism, suggesting that symptom reporting may be biased by Neuroticism-related styles of perceiving and reporting physiological experiences. Researchers in this area are urged to employ objective measures of medical status, and to be alert to possible biases of self-selection and selective perception in interpreting associations between Neuroticism and disease.