Experimental studies on chronic lymphedema of the extremities were performed using dogs, and an improved technique after the method of Danese in the preparation of chronic limb lymphedema is described. The incidence of chronic lymphedema in this model was 81.8%, whereas with Danese''s method, it was only 66.7%. In addition, the measurement of lymphatic and venous pressure was studied in experimental animals. In the normal limb, the lymphatic pressure was always lower than the venous pressure, while in the chronic lymphedematous limb, the lymphatic pressure was higher than the venous pressure. Thus, after a lymphaticovenous anastomosis, a pressure gradient reversal exists between the two sides of the anastomosis. Although the high lymphatic pressure gradually dropped as the anastomosis became patent, ultimately the venous presusre was often higher than the lymphatic pressure. A venous back flow can be ensured by the anastomosis and the resulting thrombosis, which obstructs the lymphaticovenous anastomosis, is inevitable.