Ninety-eight women with mild hyperprolactinaemia (less than 4N) were followed for a mean duration of 5.5 years. Where pregnancy was desired treatment with bromocriptine (+/- clomiphene) was effective in 87%. Following cessation of bromocriptine therapy almost one-third had a 'spontaneous' resolution of hyperprolactinaemia and resumed cyclical menstrual activity and fertility. Pituitary tumours were identified on coned-view assessment in 9% of patients at the time of presentation and a further 10% during follow-up. Although some of these latter tumours may have been diagnosed earlier had CAT scans been performed routinely in all hyperprolactinaemic patients, such a policy would be hard to justify for those with mild hyperprolactinaemia as the tumours were small, produced no harmful effects, and would almost certainly have responded to bromocriptine should this have been administered. All patients found to have pituitary tumours and given bromocriptine, showed no evidence of tumour progression subsequently, even where pregnancy occurred and the bromocriptine therapy was ceased. In addition, patients without pituitary tumours who were given bromocriptine to achieve pregnancy, were less likely to develop tumours during follow-up even when pregnancy had occurred. Bromocriptine therapy can thus be justified in mildly hyperprolactinaemic patients to reduce troublesome galactorrhoea, achieve pregnancy, improve the chance of 'spontaneous' resolution of the menstrual problem and infertility, control or reduce tumour growth where a pituitary tumour has been identified, and reduce the risk of tumour development in patients with normal radiographs at the time of presentation.