Depression is a common and serious mental illness which is treated mainly in primary care settings. The most common treatment modality is antidepressant medication. Controlled clinical trials have found antidepressants to be effective in the treatment of the acute symptoms of depression and in the prevention of relapse and recurrence of the disorder.To be effective, antidepressants need to be taken in sufficient doses for adequate periods of time. Many attempts have been made to ensure that the treatments available for depression are applied effectively in clinical practice, mainly through the development and promulgation of consensus guidelines. The various guidelines are consistent in recommending that antidepressants be administered at adequate doses both during the acute phase of treatment and for 4 to 6 months after a response is achieved.However, optimal use of antidepressants appears to be the exception rather than the rule. Patients who are treated with older tricyclic antidepressants (TCAs) rarely receive doses which have clear evidence of efficacy. Patients taking TCAs or newer antidepressants such as the selective serotonin (5-hydroxytryptamine; 5-HT) reuptake inhibitors (SSRIs) commonly stop treatment prematurely. As few as 1 patient in 17 on TCAs, and 1 in 3 on SSRIs, continues treatment at an effective dose for the minimum recommended period of 4 consecutive months.The patterns of treatment with antidepressants are completely incongruous with the magnitude of burden and risk associated with major depression. This is reflected in evidence of poor outcomes in both the short and longer terms. Failure to treat depression at an early stage contributes to the development of chronic depressive illness. Inadequate doses of antidepressants result in a failure to respond and premature discontinuation increases the risk of relapse; both these patterns of treatment seem to contribute to the development of recurrent depressive episodes.Despite these findings, suboptimal use of antidepressants in the treatment of depression appears to be universal. It seems that clinicians require clear guidance regarding the crucial importance of recognising depression and initiating effective treatment without delay. They also need information that they regard as credible regarding the vital necessity for patients to receive an effective antidepressant dose for an adequate period of time. There is compelling evidence that this is less likely to be achieved with a TCA than with an SSRI. Given the morbidity and burden caused by depression, centrally coordinated naturalistic studies to investigate the outcomes of suboptimal antidepressant use should be a key priority for government research initiatives.