A prospective study was conducted to evaluate the usefulness and limitations of conservative treatment in patients with pituitary apoplexy. Twelve patients presenting sudden headache, visual impairment, or ophthalmoplegia had the diagnosis of pituitary apoplexy established by computerized tomographic scans. Initially, 11 patients received iv dexamethasone (2.0-16.0 mg/day). Surgery was indicated when dexamethasone failed to improve visual or consciousness impairment. Among the 7 patients who were treated conservatively, ophthalmoplegia recovered completely in 6 and improved in 1. Follow-up computerized tomographic scans showed resolution of the tumor in 4 patients and residual masses in 3 patients who were treated conservatively. Five patients had surgery and experienced improvement of vision and consciousness. Follow-up computerized tomographic scans showed residual masses in all surgical patients. Recurrences were observed in 2 patients, one in each group. The prevalence of pituitary deficiencies in the conservative group (9 of 17) was similar to that of the surgical group (3 of 14), but when only patients whose tumors were resolved by the apoplexy were analyzed, a significantly higher prevalence (8 of 12) was observed (P = 0.02). A retrospective analysis of presenting clinical and computerized tomography data on the basis of the response to dexamethasone showed that visual impairment did not improve during treatment with dexamethasone, whereas the presence of a large hypodense area within the tumor predicted complete tumor resolution. These results support conservative management of pituitary apoplexy in patients who are selected on the basis of clinical and tomographic findings.