NEW MEDICINES for which a manufacturer seeks market approval are usually tested against placebo, irrespective of whether effective treatment already exists for the condition. This is a requirement set forth by the US Food and Drug Administration (FDA).1 The FDA maintains that placebo-controlled trials are necessary for an easy comparison of the new drug with "nothing," whereas evaluating therapeutic equivalence of a new vs an old medication is more problematic.2 Many patients in clinical trials are therefore deprived of effective medications. A few years ago, my colleague Ken Rothman and I drew attention to this unethical use of placebo controls.3 If effective therapy exists, the physician has an ethical and legal obligation to provide it to the patient according to ethical standards that are laid out in the Declaration of Helsinki.4 The Declaration of Helsinki mandates that a new drug be compared with the best available treatment, thus placing the interest of each individual patient above the interest of science and society. A comparison of the experimental drug with an established medication in the setting of an equivalence trial provides the patient with treatment and the physician with comparative information on safety and efficacy.