The development of surgical procedures for the removal of craniopharyngiomas (CPs) was greatly influenced by the enormous topographical and morphological heterogeneity displayed by these lesions. In this study the authors reviewed the intracranial approaches designed to treat CPs during the early historical period (1891-1938) with the aim of finding the CP topographical and pathological features that influence patient outcomes. The authors conducted a systematic retrospective review of well-described cases of surgically treated CPs in publications from the period 1891-1938. Valuable information regarding the diagnosis of the lesion, type of craniotomy performed, CP topography, and outcome was selected from 418 reports included in medical publications from this period. The type of surgical procedure used, degree of tumor removal, CP position and histological variety, and clinical evidence of postoperative hypothalamic injury were the variables analyzed with the aim of defining their influence on the final patient outcome. A collection of 160 cases was eligible for analysis. Craniopharyngioma topography was significantly related to the existence of postoperative hypothalamic damage and the degree of tumor removal achieved (p < 0.001). The infundibulo-tuberal, or not strictly intraventricular, topography was associated with the highest rate of hypothalamic injury (84%) and impossibility of tumor removal (51%). This topography also showed the worst prognosis (p = 0.001). Additional variables correlated with patient outcome were the presence of hypothalamic damage, type of surgical approach used, and degree of tumor removal. Patients having a poor outcome, suffering from permanent coma, or dying after surgery presented with symptoms of hypothalamic injury in 40% of cases (p < 0.001). The surgical approach associated with the best outcome was the transsphenoidal (58%), followed by the subfrontal (45%) and the transcallosal (45%). Subtotal resection of the lesion yielded the best postoperative results, with only 17% of patients dying or suffering from a poor outcome, in contrast to the 39% reported for gross-total removal of the lesion (p = 0.001). Two major variables influenced the results of early surgical experience with CPs for the period from 1891 to 1938: 1) the inaccuracy in defining CP topography with the diagnostic methods available at that time; and 2) the ignorance about the risks associated with the dissection of lesions showing tenacious adherence to the hypothalamus. The degree of functional and morphological disturbance of the hypothalamus caused by a CP remains a fundamental variable helping the surgeon to predict the risks associated with the radical excision of the tumor and patient outcome.