The enthusiams for injection sclerotherapy over the last decade has almost certainly surpassed what was justified on the basis of objective evidence. This was most clearly emphasized by the widespread adoption of prophylactic sclerotherapy after the report of the first two trials, even though enough was known of the natural history of variceal hemorrhage in patients with cirrhosis to warrant caution. The use of sclerotherapy for an episode of variceal hemorrhage represents the role most supported by the available data. Diagnostic endoscopy, as an integral part of management, provides the optimum time to intervene with sclerotherapy. Sclerotherapy can then provide hemostasis in patients who are actively bleeding and prevent early rebleeding in those in whom bleeding has stopped spontaneously. The progression to long-term injection sclerotherapy is of proven benefit; however, doubts exist concerning the need for the intensive regimens currently in use. The continued use of long-term injection sclerotherapy is dependent not only on additional investigations, but also on the accumulating evidence arising from comparative studies encompassing other available therapy. © 1990 Reed Publishing USA.