A Dutch family (family A) with inherited antithrombin-III deficiency and an increased incidence of venous thrombosis was investigated. Antithrombin-III levels were measured by means of a coagulation assay in plasma and by single radial immunodiffusion in plasma and serum. Three groups could be distinguished: group I comprised the relations-in-law of family A, group II the members of family A with a plasma antithrombin-III level higher than 90% of normal, when determined by the immunoassay, and group III the members of family A with an antithrombin-III level of less than 60%. To group III belonged all eight adults with an abnormal tendency to thrombosis, and furthermore nine children, all having a parent with abnormally low antithrombin-III levels. Mean plasma and serum antithrombin-III levels were significantly decreased in group III. However, the results of the coagulation assay showed some overlap between groups II and III. In addition, the immunoassay appeared to be much less laborious than the coagulation assay. Therefore, the former assay is recommended in any search for similar families. The results of our family investigation confirm the findings of Egeberg (1965) that inherited antithrombin-III deficiency, giving rise to plasma levels between 50 and 60% of normal, causes thrombophilia and that the pattern of inheritance is autosomal dominant.