MRI is a sensitive tool for the investigation of pituitary microadenomas but cannot be used as a screening investigation. To establish a strategy for the use of MRI in patients with hyperprolactinaemia we investigated 74 women with serum prolactin levels above 52 ng/ml for the presence of microadenomas. We examined 55 premenopausal and 19 postmenopausal women, using a 1.5 T unit. We used T1-weighted spin-echo sequences, with coronal and sagittal images before and after intravenous gadolinium. We found microadenomas in 38 patients (51.3%), macroadenomas in 6 (8.1%) and an infundibular glioma in 1;29 patients had a normal pituitary gland (39.2%). The size of the adenomas was related to the prolactin level, and the mean level in patients with MRI evidence of adenomas was higher than in patients without microadenomas (155.72 +/- 131.01 ng/ml versus 110.14 +/- 80.86 ng/ml). The probability of the presence of an adenoma increased with rising serum prolactin levels. We suggest MRI in patients with prolactin levels more than 100 ng/ml. In patients on oestrogen therapy MRI should be performed with only slightly elevated prolactin levels. Evidence of a microadenoma should be considered in planning further therapy, especially concerning the use of hormone replacement therapy or of bromocriptine.