A tacit assumption, which has often been made by practitioners of both chemical aversion therapy (CAT) and deterrent therapy of alcoholism, is that efficacy of a drug depends mainly on how noxious it is. We conducted two experiments which challenge this assumption. These experiments compared five noxious drugs employed in the treatment of alcoholism, with lithium, in regard to their ability to reduce rats' drinking of a familiar, readily consumed, ethanol (10% v/v) solution. In both experiments, the dose of lithium was equivalent, in terms of body weight, to the largest daily recommended human dose: the dose of each of the five, noxious drugs was twice this equivalent. In the first experiment, the CAT drugs emetine, ipecac and succinylcholine were compared with lithium: in the second, the deterrent therapy drugs citrated calcium carbimide and disulfiram were compared. Despite the use of comparatively smaller doses of lithium, none of the noxious drugs administered in either experiment produced stronger ethanol aversions than lithium. This supports the use of lithium in CAT, and suggests that unnecessarily noxious drugs are being used in the treatment of alcoholism.