The large number of variables in any clinical problem demands careful organization of an investigative program to yield meaningful results. The evaluation of different therapy schedules and plans in superior vena caval obstruction patients is particularly vexing. The neoplastic etiology may range from exquisitely sensitive lymphomas to moderately responsive well differentiated bronchogenic cancers. The onset may vary from the acute and dramatic to the slow and chronic, with different degrees of collateral circulation and edema. Philosophies as to the use of irradiation alone or combined with chemotherapeutic agents run the gamut of clinical opinion, depending upon the radiotherapist, the internist, and the thoracic surgeon consulting on the case. The emergency nature of the presentation of acute caval obstruction requires a consistent and well defined approach to effect rapid relief of symptoms. The varied opinions as to proper therapy coupled with the relatively small numbers of cases led to the search for a biological model to simulate the clinical problem. After a number of attempts with different animals and tumors, the successful experimental production of the superior vena caval obstruction syndrome was accomplished. A number of different studies were designed to answer specific problems, and the results of this program have served as a guide for modifying our clinical treatment plan. Method of Study Sprague-Dawley female rats of approximately 150-gram weight were the animals used. The tumor utilized was the Rochester strain of the Murphy-Sturm lymphosarcoma, a tumor of rapid growth and radioresponsiveness. Although termed a lymphoma, its histologic characteristics simulate very anaplastic carcinomas. Donor rats with approximately seven- to ten-day old flank tumors were sacrificed. Healthy-appearing sections of tumor were morcellated in isotonic saline; a small amount of this mixture was then implanted percutaneously under the superior sternum with a No. 15 trocar (Fig. 1). Other rats from the same group underwent implantation in the flank and served as controls. Ether anesthesia was used during these procedures. The animals with mediastinal tumor implants were followed daily by supine chest radiographs and clinical observation. When the superior mediastinal shadow had widened 3 mm. or more and the clinical picture of an acute superior vena caval obstruction developed, various therapeutic measures were undertaken. Observation continued both by chest radiograph and by clinical observation. Each separate group studied included 10 to 15 animals. Without exception, the animals all survived the implant and appeared normal for at least the next forty-eight to seventy-two hours. Tumor takes were greater than 90 per cent in all groups studied. In this particular strain, spontaneous tumor regression occurred in only 1 or 2 per cent of the animals.