OBJECTIVE Clinicians have difficulty interpreting results of tests for anti-cardiolipin antibodies (aCL) because of conflicting reports of the clinical associations of these antibodies in patients with systemic lupus erythematosus (SLE). We therefore decided to evaluate the clinical associations of aCL in an effort to facilitate interpretation of single reports of either positive or negative test results. We also assessed the role of estrogen on the development of aCL. METHODS The study population consisted of 85 consecutive outpatients with SLE and 40 control subjects. Serum samples and clinical and laboratory data were obtained from each patient and control. Testing for aCL was performed using a standardized enzyme-linked immunoabsorbent assay developed at an international workshop. RESULTS The presence of aCL was documented in 42.4 percent of patients with SLE and 7.5 percent of control subjects. In patients with SLE, these antibodies were significantly associated with thrombosis, fetal loss, and thrombocytopenia, but not with other manifestations. Measurement of all isotypes optimized clinical correlations. Titers did not add clinical utility. Fluctuations of levels of aCL occurred, making it difficult to interpret a single negative result. Among control subjects, the presence of aCL was not significantly more common in women who used oral contraceptives. CONCLUSIONS Our findings suggest that positive results of testing for aCL correlate with a predisposition for thrombosis, fetal loss, and thrombocytopenia in patients with SLE; however, the test is not predictive for other clinical manifestations of SLE, including activity and severity of disease. We believe that measurement of all isotypes of aCL should be performed in patients with SLE considering pregnancy, to identify those with a high risk of fetal loss, and in SLE patients with a thrombotic episode.