In addition to posing a risk for coronary heart disease (CHD), Type A behavior has been suggested as a risk factor for non-CHD illness. Past evidence, however, has relied chiefly on retrospective studies of self-reported illness that failed to control for the potentially confounding effects of negative affectivity. The present investigation was an 18-month prospective study of Type A behavior and medical records of illness in which chronic negative affectivity was also assessed. Negative affectivity was associated with retrospective self-reported illness, but only Type A, as measured by the Jenkins Activity Survey (JAS) and the Framingham Type A Scale (FTAS), predicted medical records of subsequent illness severity and number of illness-related medical visits. Controlling for negative affectivity did not reduce these significant relationships. It was concluded that persons scoring high on the JAS or the FT AS may be at greater risk of minor illness when objective measures of illness are assessed over a period of 1 year or more. Alternative explanations and future directions for research are discussed.