Throughout history, doctor-patient relationships have been acknowledged as having an important therapeutic effect, irrespective of any prescribed drug or treatment. We did a systematic review to determine whether there was any empirical evidence to support this theory. A comprehensive search strategy was developed to include 11 medical, psychological, and sociological electronic databases. The quality of eligible trials was objectively assessed by two reviewers, and the type of non-treatment care given in each trial was categorised as cognitive or emotional. Cognitive care aims to influence patients' expectations about the illness or the treatment, whereas emotional care refers to the style of the consultation (eg, warm, empathic), and aims to reduce negative feelings such as anxiety and fear. We identified 25 eligible randomised controlled trials. 19 examined the effects of influencing patients' expectations about treatment, half of which found significant effects. None of the studies examined the effects of emotional care alone, but four trials assessed a combination of both cognitive and emotional care. Three of these studies showed that enhancing patients' expectations through positive information about the treatment or the illness, while providing support or reassurance, significantly influenced health outcomes. There is much inconsistency regarding emotional and cognitive care, although one relatively consistent finding is that physicians who adopt a warm, friendly, and reassuring manner are more effective than those who keep consultations formal and do not offer reassurance.