Abstract
Data on the 67 women asked by this study and by their physician about their drinking indicate that most of them reported "occasional" consumption to the physician, no matter what their drinking behavior appeared to be as estimated by personal interview. That is, heavy drinkers (by the estimates of this study) underestimated and infrequent drinkers overestimated alcohol consumption relative to the personal interview when reporting to their physician. Becuase of the difficulty in classifying responses, a rather loose criterion of "disagreement" of the two estimates was established. About 10% of the women gave conflicting reports to the study interviewer and the physician by this criterion. In most cases of conflict, the report to the physician was lower. These data suggest that the reports of drinking to the physician were of limited usefulness, becuase all women tended to declare similar quantities. Furthermore, physician interrogation failed to detect almost all the heavier drinkers found by an independent interview.