Placebo effects are often attributed to clinical interactions and contextual factors that affect expectations of the patient about the treatment and result in symptom changes. The prevailing conceptualization consists of an undifferentiated placebo response that needs to be minimized in controlled investigations and maximized in clinical practice. However, treatment outcome is the cumulative result of the interaction of several classes of variables with a selected treatment: living conditions (housing, nutrition, work environment, social support), patient characteristics (age, sex, genetics, general health conditions, personality, well-being), illness features and previous therapeutic experience, self-management, and treatment setting (physician's attitude and attention, illness behavior). Such variables may be therapeutic or countertherapeutic, and are unlikely to be simply additive. In certain patients their interactive combination may lead to clinical improvement, whereas in other cases it may produce no effect, and, in a third group, it may lead to worsening of the condition. Maximizing patients' expectations does not necessarily result in sustained effects and, in due course, may actually lead to worsening of the condition (violation of expectations). In this paper, we outline a multifactorial conceptual model that may have implications for the design of clinical trials as well as for clinical practice, with special reference to psychopharmacology and psychotherapy. The effects of drug treatment may be potentiated by specific nonpharmacological treatment strategies, and this synergism may disclose significant differences against placebo. Medical outcomes may be unsatisfactory not because technical interventions are missing, but because our conceptual models and thinking are inadequate.