It is said that early work on the biological faise positive seroiogical test for syphilis was stimulated by the indignant reaction of an immunoiogist's wife to the suggestion that she had a venereal disease rather than the connective tissue disorder that she later developed. It was not just his wife that protested. The test for syphilis originally described by Wasserman [1] in t907, based on antibodies to a reagent eventuaiiy characterized as cardioiipin in 194i [2], had been compuisory for ati candidates for marriage and pregnant women, required by iaw in 40 American states. After the iast War, over 75 000 US services personnei were found to have seroconverted-and over haif of these were put down to biological faise positive reactions. At that time it was estimated that the incidence in systemic iupus erythematosus (SLE) was 20% (and 5% in rheumatoid arthritis), but there were also other diseases with a high incidence, particularly infections such as maiaria (100%) [3].These data, and the hypotheses arising from them, have been regularly revisited, not ieast in this journal. There are 'natural' antibodies to cardioiipins in normal human subjects [4], and a B cell clone from a patient with chronic iymphocytic leukaemia (CLL) has been demonstrated to make pathogenic anticardiolipin antibody using genes in germ-iine configuration [5]. These may be analogous to the very common reactivity, using restricted idiotypes, to simple haptens iike dinitrophenyi in the normai mammalian B cell repertoire. However, there is no doubt that a number of exogenous antigens can stimulate these antibodies, particularly infectious agents [6]. In this issue of Clinical and Experimental Immunology, another group [7] has shown that antibodies can bind to cardioiipin in association with co-factors (histones in this study), and the authors link this finding to the known antibody reactivity to histones/DNA in connective tissue diseases, and postulate that this may be diagnostically if not pathogenicaiiy important.What are these antigens and co-factors, and what specificity do the antibodies have in common? The recognition of the simitar anionic phosphotipid nature of the laboratory reagents used in the diagnosis of syphilis and the lupus anticoagulant, and the more recently available purified cardiotipins, seemed insufficient on its own to explain the cross-reactive binding. Co-factors have now been described which ciarify many of the earlier problems with the assays for antibodies to phosphotipids and their relevance in disease.The first distinction made was between the 'acute' and the 'chronic' antibodies [t]. The acute antibody reactions to