Venous thrombosis is an infrequent but serious cause of hospitalization in children. The epidemiology and natural history remains incompletely defined, especially in geographically distinct regions of the United States. We thus evaluated thrombosis in a single children's hospital over a 3-year period. Of 41,906 hospitalizations, 92 children were identified for review. The incidence of thrombosis was 21.9 per 10,000 admissions (0.22%). Venous thrombosis was of equal incidence in African-American and white patients. Locations of thrombosis included deep venous (51%), pulmonary (21%), renal vein (8%), intrahepatic (8%), and intracranial (12%). Risk factors for thrombosis included central catheter (32%), malignancy (18%), systemic infection (21%), neurologic disability (9%), cardiac (4%), nephrotic syndrome (3%), and autoimmune (6%). Six of 92 patients (7%) had thrombophilia. Positive family history of venous thromboembolism (VTE) or thrombophilic disorder predicted an abnormal test. Treatment included low-molecular-weight heparin (n=53), coumadin (n=12), heparin (n=10), tissue plasminogen activator (n=6), argatroban (n=1), thrombectomy (n=2), inferior vena cava filter (n=2), and no treatment (n=23). Seventy-seven percent demonstrated resolution of the VTE, 14% had persistent or recurrent VTE, and 9% died. Causes of death were malignancy, prematurity, septicemia, and congenital heart disease. Venous thrombosis is a serious comorbidity in hospitalized children. In our population, African-Americans had an equal incidence of VTE as whites. Positive family history predicted thrombophilia.