During the last decade several studies have indicated that sleep problems might belong to the risk indicators for coronary artery disease (CAD). For example, a 6-year follow-up study of >10,000 subjects revealed a risk of 2.04 for CAD in "poor" versus "good" sleepers.1 Little is known about the types of sleep complaints that are associated with future CAD. One may speculate that trouble falling asleep is associated with CAD, because this is indicative of prolonged tension. Trouble staying asleep may be predictive for the same reason, or because it is indicative of a heart failure or nocturnal angina. Waking up tired may reflect an impaired sleep or an adverse effect of medication, angina pectoris or aging, and lose its predictive power when adjusted for these factors. Waking up tired may also indicate depression. Recent meta-analyses of the vast literature of personality factors and CAD have shown that, of all personality attributes, depression is the one most strongly associated with disease outcome.2 Because early-morning tiredness is a major characteristic of depression it is predicted that problems falling asleep or staying asleep are not predictive of MI when controlled for waking up exhausted, but that the latter complaint remains predictive when controlled for complaints indicating problems in gaining or maintaining sleep (hypothesis I). Waking up exhausted may be a consequence of poor sleep. Statistically controlling for the influence of problems falling or staying asleep does not completely remove any possible effects of these problems on the feeling of being tired at waking up. To rule out any possibility that exhaustion on waking up is caused by a "bad night," we also tested the hypothesis that those who wake up exhausted but do not have problems falling or staying asleep are at increased risk of myocardial infarction (hypothesis II). © 1991.