Recently published encouraging data concerning the value of anticardiolipin antibody assays in patients with HIV infection, both for prognostic purposes and to diagnose pneumocystosis, have prompted us to evaluate these assays in a series of 116 patients. We were unable to demonstrate any correlation between the levels of these antibodies and the clinical stage of the infection or the presence of thrombocytopenia. Varying levels of anticardiolipin antibodies were found in 29.3 percent of our patients. More frequent positivity and higher levels seemed to be associated only with the development of cerebral toxoplasmosis. The value and the possible physiopathological role of these antibodies in patients with HIV infection should be reconsidered.