BACKGROUND: A number of surgical series have been reported on the treatment of acromegaly and their results vary widely. The acceptable definition of remission has changed in recent years and it is known, though in a small series, that growth hormone levels of > 5 mU/l are still associated with an increased mortality from the condition. We have analysed data at this centre and examined the outcome of transphenoidal surgery for acromegaly, compared our results with recently published series from other centres and also assessed factors which might effect outcome including whether there is any demonstrable effect of the experience of the surgeon on outcome. PATIENTS AND METHODS: We have analysed data from all of our 139 patients in whom follow up data are available who have undergone initial transphenoidal surgery for acromegaly by one surgeon at this centre, between 1974 and 1995. Follow up was for a median of 5 years (range 1 month to 17 years). RESULTS: 67% of patients achieved the criterion for remission (mean GH < 5 mU/l). Success was related to tumour size and preoperative growth hormone values. Thus 91% of patients with microadenomas were in remission postoperatively compared to 46% of patients with macroadenomas. Analysis of the results according to the year of operation showed an improvement in success rates with time. More than 15 years ago, the success rate according to the growth hormone criteria set was 48% and the failure rate 52%. In contrast in the last 5 years analysed, the overall success rate was 74% with a failure rate of 26% (P < 0.04). The success rate for microadenomas was 50% initially, then remained 100%. The case mix was analysed and no change was found. We have also demonstrated an improvement in pituitary function (including normalization of preoperative elevated prolactin) with time so that pre 1981 16% of patients' pituitary function improved perioperatively but 10 years later this figure had risen to 34% (P < 0.03). There was no change over time in the development of pituitary hypofunction, complication rate or recurrence rate. CONCLUSION: Surgical treatment is a safe and effective treatment for acromegaly and remains the first choice of treatment for most acromegalic patients. The results of this centre compare favourably with series from other centres. We have demonstrated improved results, both in terms of post operative growth hormone values and pituitary function tests with time and increasing neurosurgical experience. We conclude that outcome for the surgical treatment for acromegaly is best achieved with one surgeon specialising in pituitary surgery. Improved operative outcome thus achieved has major cost implications and avoids the necessity for consideration of postoperative radiotherapy and the use of expensive growth hormone suppressing drugs in the postoperative period.