In a series of approximately 1,000 patients with pituitary adenoma who were operated on at our institution from 1970 to 1994, 932 were operated on for the first time. Most microadenomas were corticotropin (ACTH)-secreting, whereas all nonfunctioning adenomas were macroadenomas, reflecting internal surgical policy. Only 48 of 932 patients (5.1%) had transcranial surgery. Using stringent criteria for the definition of a cure, we obtained remission of disease in 54.9%, 87.3%, 66.7%, and 46.2% of all patients with growth hormone (GH)-, ACTH-, thyrotropin (TSH)-, and prolactin (PRL)-secreting adenomas, respectively. The good result in patients with Cushing's disease is related to the higher percentage of microadenomas (approximately equal to 80%) in this group. Unfavorable prognostic characteristics for all adenomas are increasing tumor size, invasiveness, infiltration, and high serum levels of the hypersecreted hormone. The absence of a reliable tumor marker makes it difficult to assess the results of surgery in patients with nonfunctioning pituitary adenoma, but normalization or improvement of visual defects occurred in 72.4% of patients. Permanent worsening of vision was detected in 2.2%, mostly operated on through the transcranial approach, but they had large tumors and so were at greater risk. Accordingly, there was a higher death rate in patients who received the transcranial operation (two of 48; 4.2%) than in patients operated on through the trans-sphenoidal route (seven of 884; 0.8%). However, between 1970 and 1980, the mortality rate was 1.6% (six of 367 patients), while between 1981 and 1994, it was 0.5% (three of 565 patients), stressing the importance of surgical experience and perioperative medical management in improving the safety of pituitary surgery.