Two cases are presented of soldiers who suffered from recurrent, disabling, hard swelling on the dorsum of the hand, following a severe initial blow. In each case operative exploration and examination of pathological specimens revealed evidence of old and recent hemorrhage, with fibrous-tissue proliferation and organization in the hemorrhagic area. Infiltration of the extensor tendon by the fibrous tissue gives explanation for the clinical finding of local pain and limitation of movement during flexion of the fingers: also, it is evident that, with extremes of motion of the tendon, disruption of the attached fibrous tissue occurs with further hemorrhage, clotting, fibroblastic organization, and increased fibrosis deep to the superficial fascia. This explains the recurrent, localized, and hard character of the swelling. In such cases it would appear that the treatment of choice is early operative evacuation of the primary hematoma and ligation of any evident bleeding vessels, followed by a firm compression dressing. Involvement of the extensor tendons during the fibroblastic organization of the adjacent hematoma cannot otherwise be avoided with certainty. In late cases, evacuation of the hematoma should probably be followed by prolonged immobilization, in order to allow maturation of the fibrous tissue and obliteration of the hematoma cavity. The authors are not prepared as yet to state the value in advanced cases of excision of the involved tendons, subcutaneous tissue, and skin, with replacement by a pedicle skin graft and tendon transplants. Such procedures are certainly contra-indicated in cases with a factitial etiology, in old cases with severe joint contracture and marked loss of function, and in those with marked secondary sympathetic-nerve phenomena. Probably not all cases of peritendinous fibrosis of the dorsum of the hand have a similar basis. In our opinion this present hemorrhagic type constitutes a definite subdivision of a group of ill-defined cases which follow a definite syndrome and are resistant to the usual treatment. The presence or absence of this hemorrhagic type should always be established before other operative procedures, such as various types of cervical sympathectomies, are performed. Permanent good results cannot logically be anticipated with cervical sympathetic procedures in those cases in which fibroblastic infiltration and fixation of the extensor tendons have occurred.