Data from 9740 Scottish men and women aged 40-59 years, selected at random in a population survey during 1984-86, were used to compare self-reported smoking habits and biochemical measures of tobacco smoke inhalation in three groups: 625 with diagnosed coronary heart disease; 1497 undiagnosed, but with suggestive questionnaire symptoms or electro-cardiographic signs; and 7618 with no diagnosis or evidence of coronary disease, nor cardiovascular medication. Men and women with a diagnosis of coronary heart disease reported that they had tried to reduce their cigarette consumption or had quit more often than did the two comparison groups, but this effect was considerably greater in men than in women. Men in all groups had a higher cigarette quitting rate, but this may be due to switching to pipes and cigars. However, the biochemical measurement of tobacco smoke inhalation by cotinine showed that those with a coronary diagnosis had the highest overall levels of tobacco smoke inhalation, and the highest level of cotinine per self-reported cigarette, cigar, or ounce (30 g) of pipe tobacco consumed. Self-reporting after a coronary diagnosis is consequently distorted, and an objective measure of smoke inhalation is to be preferred. In both sexes, particularly women, intake of tobacco products remains higher than that in the general population, whether or not it was even higher before they were diagnosed. Those responsible for continuing care of coronary patients need to be aware that health education and compliance are less effective in this group than they should be, both in male patients and, even more so, in women.