Like a lot of medication, lithium was introduced through a mistaken hypothesis (see webpage). I'm not convinced its efficacy has ever been satisfactorily demonstrated. Rapid withdrawal is associated with higher relapse than gradual withdrawal (see my eletter). Such a finding suggests that nonspecific factors are important in the interpretation of clinical trials (see another eletter). Unblinding can occur in clinical trials of lithium, compromising their results (see my BJPsych letter). Denial of the extent to which lithium may be a placebo effect in clinical practice does not necessarily serve the interests of the many patients who have been made reliant on this medication.