The monoamine hypothesis has dominated our understanding of depression and of pharmacological approaches to its management and it has produced several generations of antidepressant agents, ranging from the monoamine oxidase inhibitors (MAOIs), through tricyclics (TCAs) and selective serotonin reuptake inhibitors (SSRIs), to the recently introduced selective noradrenaline reuptake inhibitor (NARI), reboxetine. Greater receptor selectivity has improved tolerability, but not efficacy, when newer compounds are compared with the original tricyclic antidepressants (TCAs) and monoamine oxidase inhibitors. Essentially, the newer antidepressants have the same distinguishing feature as older ones, i.e. acute enhancement of monoaminergic neurotransmission. The monoamine hypothesis cannot conclusively link the acute biochemical action of antidepressants on monoamine levels with their delayed clinical effect of 10–14 days, nor can it explain the mode of action of antidepressants that are effective despite being very weak inhibitors of monoaminergic transmission (e.g. iprindole) or, incongruously, enhancing monoamine uptake (e.g. tianeptine). Compared with other fields of medicine, there has been a lack of progress in understanding the pathophysiology of depression and producing truly novel antidepressant agents. Other biological approaches to depression, such as overactivity of the hypothalamic–pituitary–adrenal axis, hippocampal neural plasticity in response to stress, and the link between the inflammatory response and depression, offer new approaches to finding pharmacological agents, aided by improved techniques for visualising the human brain, better animal models, and increased knowledge of human markers of depression. Copyright © 2001 John Wiley & Sons, Ltd.