We retrospectively analysed the long-term treatment results (median 8 years) of 31 patients with macroprolactinoma. 17 patients were treated by pituitary surgery (group 1) followed by long-term dopamine agonist therapy whereas 14 patients received long-term dopamine agonist therapy alone (group 2). 2 patients of group 1 and 1 patient of group 2 had external pituitary irradiation because of progressive disease. The two groups were comparable with respect to age, gender and initial prolactin (PRL) levels. At the end of the observation period dopamine agonist dosage could be reduced by 50% in group 1 and by 39.3% in group 2. Pituitary function did not change substantially during therapy. Complete remissions (no visible tumour in CT or MRI, normal PRL levels under current dopamine agonist medication) were achieved in 23.5% of group 1 vs. 21.4% of group 2, partial remissions (reduction of PRL and tumour size) in 35.3% vs. 64.3%, stable disease in 23.5% vs. 7.1% and progressive disease in 17.7% vs. 7.1% (differences not significant). Visual field defects showed 28.4% remissions (complete and partial) in group 1 versus 50% in group 2. Dopamine agonist therapy could be stopped definitively in only 1 patient of group 2 with an invasive macroprolactinoma. Initial surgical reduction of tumour load followed by medical therapy does not seem to guarantee a better long-term outcome than dopamine agonist therapy alone if the patient responds to and tolerates dopamine agonist therapy. Surgery should be reserved for non-responders, drug-intolerant or non-compliant patients, and for those with acute severe neurological compromise.