Until about thirty years ago, the extent of disclosure about and consent-seeking for medical interventions was influenced by a beneficence model of professional behaviour. Informed consent shifted attention to a duty to respect the autonomy of patients. The new requirement arrived on the American scene in two separate contexts: for daily practice in 1957, and for clinical study in 1966. A confusing double standard has been established. 'Daily consent' is reviewed, if at all, only in retrospect. Doctors are merely exhorted (...) to obtain informed consent; they often minimise uncertainties about 'best' treatment and they feel duty-bound to provide patients with an unequivocal recommendation for action. 'Study consent' in a clinical trial is reviewed prospectively, and doctors are compelled by regulation to point out that there is insufficient evidence to make a rational choice between two compared treatments. It has been impossible to devise informed consent practices that satisfy, in full, the competing moral imperatives of respect for autonomy, concern for beneficence with emphasis on the value of health, and a vigil for justice. A way must be found to experiment with various discretionary approaches that would strike a realistic balance among competing interests. (shrink)