Unfortunately there is still a marked tendency on the part of many surgeons to overlook completely, or ignore until too late, those signs and symptoms which are indications for surgery in injuries of the spine and spinal cord. All too frequently the general attitude is that a cord injury is an utterly problem, and it is useless to do anything about it. As a result, those individuals who might have had a chance of recovery actually do become hopeless cases. The early indications for surgery in spinal-cord injuries are usually given as: 1. A history of progression of neurological signs. 2. Compound fractures of the spine, particularly in those cases with spinal-fluid leaks. 3. Marked encroachment of bony fragments upon the spinal canal, as exhibited in the roentgenograms. 4. With a lumbar-puncture needle in place, evidence of complete blockage of the spinal fluid on jugular compression. To these it is believed a fifth should be added: 5. Evidence of blockage of a column of pantopaque upon myelographic examination in those cases which, in spite of a negative Queckenstedt test, have a definite sensory level. The first two indications listed require no discussion, but the last three should be commented upon. It should be emphasized that, at operation, 30 or 40 per cent. more damage to the vertebrae may be noted than can be visualized in the roentgenogram. As shown in Case 2, if one relies merely upon a positive jugular-compression test as an indication that operation should be performed, extruded discs may be missed. It is obvious that this test will be negative as long as a subarachnoid space of the diameter of a No. 20 needle is still present. The validity of the test is good for that moment during which the procedure is being performed. It is known that oedema may gradually develop in the injured cord or cauda equina hours or even a few days after this initial test has been performed; and, unless the procedure is repeated several times, a true surgical lesion may be overlooked. For this reason it is advocated that early pantopaque myelography be performed in these lesions where there is no block to jugular compression, and yet a good sensory level is present. A word of caution is offered to those doing myelography on these individuals. In order to prevent further damage to the injured cord, the patient should be anchored securely to a board on the x-ray tilt table, with a firm supporting yolk at the shoulders and adjustable ankle supports. This enables tilting of the patient into an almost vertical head-down position, without compression of the vertebral column or the spinal cord. It would seem unnecessary to emphasize these criteria, but only too rarely are they followed. In all of the cases reported here, operation was performed much later than it should have been, and the degree of disability has, at least theoretically, been increased proportionately. These cases should be regarded as surgical emergencies. After early accurate diagnosis, operation should be performed immediately by a competent surgeon, for the elements of the central nervous system do not, in general, withstand prolonged compression without the occurrence of irreversible anatomical changes.