Newer antidepressants, such as the selective serotonin reuptake inhibitors (SSRIs), selective noradrenaline reuptake inhibitors (NRIs) and drugs acting on both serotonin and noradrenaline, represent a clinically significant advance over the tricyclic antidepressants (TCAs) and monoamine oxidase inhibitors. Nevertheless, all current pharmacological treatments for depression fall somewhat short of the 'ideal', and unmet needs remain in the pharmacological management of depression. In general, current antidepressants differ little in terms of efficacy. However, in terms of tolerability, SSRIs appear to offer clear advantages over older therapies. For example, the proportion of patients who discontinue treatment because of adverse events and the risk of fatal toxicity both appear to be lower with SSRIs, compared with TCAs. Some adverse events, however, such as sexual dysfunction and nausea (typical serotonergic side effects), tend to be more frequent with SSRIs. Also, individual SSRIs appear to differ in the likelihood that they will be associated with these side effects. However, no clear advantage emerges for any one SSRI over another, and the current recommendation is that the anticipated side-effect profile should be tailored to the patient. Current pharmacoeconomic data do not appear to favour any antidepressant over another, with no cost benefits associated with prescribing TCAs instead of SSRIs.The refinement of current drugs arising from the introduction of their active enantiomers may translate into clinical benefit, bringing closer the 'ideal' antidepressant. Potential benefits of a single enantiomer versus the racemate in the management of depression include: a reduction in the total dose, while maintaining desired outcomes, as well as a possible greater efficacy, dose for dose,simpler assessment of dose-response relationships,a reduction in pharmacokinetic and pharmacodynamic variability between patients, anda reduction in any toxicity arising from the inactive stereoisomer. However, these benefits remain to be proved in clinical trials and naturalistic studies. Further refinement of the benefit:risk ratio of current 'gold standard' drugs may go some way to addressing the shortfalls in existing antidepressant therapy. Copyright 2001 John Wiley & Sons, Ltd.