We describe here a test for lupus anticoagulants based on a modified Russell viper venom time (RVVT), using limiting amounts of phospholipid and venom. We have studied 29 patients with a prolonged dilute RVVT. Five of the 29 had a normal activated partial thromboplastin time and three of 14 tested by the tissue thromboplastin inhibition test were normal. In 17 of 19 patients tested, the dilute RVVT was completely normal when ionophore-treated platelets were substituted for phospholipid; the remaining two patients, both with very long phospholipid-dependent dilute RVVT's, were nearly completely normalized. The dilute RVVT is not prolonged in the presence of antibodies to factors VIII, IX, or XI. Thus, the dilute RVVT appears to be a simple, reproducible, sensitive, and relatively specific method for the detection of lupus anticoagulants.