Objective: Dopamine agonists are the primary therapeutic modality for the majority of patients with prolactinomas, with pituitary surgery reserved for those patients intolerant of or resistant to these agents. Most published surgical series, however, contain patients treated by surgery as the primary therapeutic modality. Previous exposure to dopamine agonists or the selection of patients with prolactinomas resistant to conventional therapy may potentially compromise the surgical success rate. The purpose of this study was to evaluate the efficacy and safety of pituitary surgery for prolactinomas in a tertiary referral centre where the majority of patients were operated on after treatment with dopamine agonists. Design: A retrospective review of the outcome of pituitary surgery for prolactinomas performed at a tertiary neurosurgical centre by a single neurosurgeon. Patients: Twenty-three patients underwent excision of a macro and 11 excision of a micro-prolactinoma. Measurements: Pituitary tumour diameter was determined by CT or MRI imaging. Pre and post-operative measurements were made of serum PRL concentration (off dopamine agonist therapy), free T4, free T3, LH and testosterone (males). Post-operative restoration of a menstrual cycle was taken to indicate resolution of hypogonadism in female patients. Results: The majority (73.9%) of the patients with macro and all with micro-prolactinomas had received dopamine agonists preoperatively. Of the 23 patients with macroprolactinomas, in whom the median preoperative PRL concentration was 13255 mU/l, 17 (73.9%) had radiological evidence of suprasellar extension and 5 (21.7%) cavernous sinus invasion. Only 4 (17.4%) of the patients with macroprolactinomas had a normal serum PRL post-operatively, although there was an improvement in visual fields in 66% of those with preoperative defects. The median preoperative PRL concentration was 4309 mU/l in the patients with microprolactinomas, significantly lower than in the macroprolactinoma group (P = 0.02). Despite a significant fail in serum PRL postoperatively (median PRL 860 mU/l, P = 0.0001), only 45.5% of patients had a normal serum PRL concentration after surgery. Conclusions: The cure rate following pituitary surgery for prolactinomas in a tertiary referral centre was low when compared with previous series in which surgery was used as the primary therapeutic modality. We suggest this may result both from dopamine agonist pretreatment and the referral of prolactinomas resistant to conventional therapy. The outcome is probably a more realistic reflection of the results of pituitary surgery for prolactinomas as currently practised in the majority of neuroendocrine centres.