Seventy-nine patients with acromegaly were investigated before and after transsphenoidal adenomectomy, to determine the immediate and late outcome, the pre-operative features associated with a good result, and the accuracy of post-operative testing in predicting outcome. Pre-operative evaluation included basal growth hormone (GH), GH response to oral glucose tolerance test (OGTT), GH response to thyrotrophin-releasing hormone (TRH), tests of pituitary reserve, and pituitary scanning to assess tumour size. A few weeks after surgery, these tests were repeated. The patients were recalled for late assessment 1-13 years (median 86 months) after the operation. At the immediate postoperative testing, minimum GH after oral glucose was < or = 2 mU/l in 48.7%, < 5 mU/l in 76.3% and < 10 mU/l in 84.2%. Only 12 patients had GH > 10 mU/l. Basal GH was < or = 2 mU/l in 21%, < 5 in 59.2%, < 10 in 73.6% and < 20 in 90.8%. A minimum GH of < or = 2 mU/l during an OGTT was achieved in 67.4% of patients with intrasellar tumours, compared with 27.3% with extrasellar tumours. Basal GH and post-glucose GH correlated with the late outcome. GH response to TRH showed no correlation with outcome. IGF-1, which could not be assessed in detail, correlated with GH but was not a reliable indicator of outcome. Transsphenoidal adenomectomy is thus a very satisfactory treatment for acromegaly. Postoperative levels of basal growth hormone < 5 mU/l and post-glucose GH < or = 2 mU/l can be regarded as a biochemical cure. Postoperative radiotherapy is not required in patients who achieve a good result. The preoperative factors which significantly influenced the final outcome were basal GH, post-glucose minimum GH, tumour size and impaired pituitary reserve.