Acutely administered antidepressants possess a multiplicity of pharmacological actions. However, the fact that agents possessing similar pharmacological actions are devoid of antidepressant activity, together with the lack of correlation between doses required for acute pharmacological effects and clinical efficacy, suggest that the mechanism(s) of action of antidepressants cannot be directly attributed to the acute pharmacological properties of the drugs. The lag phase in onset of clinical effectiveness emphasizes the importance of adaptive changes following chronic antidepressant administration. A rapidly accelerating trend in attempting to delineate the precise molecular mechanisms of action of antidepressants is the shift in emphasis following chronic antidepressant therapies from alterations in uptake, storage, synthesis and release of neurotransmitters to adaptive changes in receptor functioning. These adaptations occur both pre- and postsynaptically. Examples of the former are alpha 2 and DA presynaptic receptors, both being down-regulated by certain forms of chronic antidepressant therapy. The fact that the NE-coupled adenylate cyclase system in rat brain slices is down-regulated by tricyclics, atypical antidepressants, MAO inhibitors and ECT emphasizes the importance of the system. Electrophysiological and behavioral studies point to the up-regulation of central alpha 1 and 5-HT receptor functioning following long-term antidepressant therapy. In contrast to the beta-adrenoceptor, these findings cannot be correlated with data from radioligand binding studies. In general central alpha 1-adrenoceptor binding remains unaltered. This is also true for 5-HT1 binding whereas cortical 5-HT2 binding is both increased and decreased depending on the type of antidepressant therapy being investigated. The relationship of these adaptive changes to the clinical efficacy of antidepressants in man is not clear since there is generally a lack of good models for studying human central receptor functioning. A review of current data from animal studies would tend to disfavour the view that all forms of antidepressant therapy possess a common mechanism of action. Perhaps multiple intervention sites exist. The introduction and evaluation of agents possessing a specificity of pharmacological action will undoubtedly aid psychotherapeutic research. The knowledge that peptides and 'classical' neurotransmitters can co-exist in the same neurone will undoubtedly generate studies of the significance and importance of the co-transmitter function of peptides in the mechanisms of action of antidepressant therapies.