The process of fertilization in humans, is remarkably inefficient. Spontaneous abortion is estimated to be between 15 and 20% of all clinical pregnancies, and the early spontaneous abortion rate is closer to 30-50% of fertilized ova. Not all authors agree on the definition of "recurrent spontaneous abortion" (RSA), so the frequency of repeated pregnancy wastage is difficult to determine; from empirically derived data, it has been estimated to range between 0.4 and 0.8%. Because of the various etiologies of RSA, their association in determining an abortive event, it is difficult to evaluate their exact incidence. Moreover, their is no prospective study on this subject, so it is advisable to distinguish between the admitted causes, the likely factors, and the etiologies to be evaluated. In the first group, the congenital or acquired müllerian anomalies (especially the septate uterus), represent about 25% of the RSA, but a lot of problems concerning the physiopathology are still debated, even if the rate of pregnancies after surgery ranges around 50% in certain series. On the other hand, the genetic factors, identified especially with the banding technique, are undeniable: however, although the rate of chromosomal aberrations in the offspring (Monosomy X, Trisony 16, Triploidy) is very high (50 to 60% of spontaneous abortions in the first trimester of pregnancy), when couples with usual abortions are subjected to karyotypic analysis, genetic anomalies (especially translocations) are been noted in only 6.2% of the women and 2.6% of the men. In the second group, the infective factors (chlamydiae, toxoplasma and mycoplasma) are difficult to analyse since the serology is not sufficient without a real proof of an endometrial colonization. Among the endocrinological causes, the classical luteal phase deficiency remains a subject of controversy (estimated between 3 and 30%) not only for the establishment of the diagnosis, but also for the efficiency of progesterone supplementation. In the third group, the autoimmune diseases, especially systemic lupus erythematosus (SLE) are most likely a cause of fetal wastage, by secreting antibodies that cross-react with the antigenic determinants of the trophoblast. These patients should be screened for lupus like anticoagulant; the therapy with glucocorticoids during pregnancy is still debated. More recently, the immunologic factors have been estimated to play an important role in the etiology of usual abortion: in these cases women share HLA antigens with their partners more frequently than expected.(ABSTRACT TRUNCATED AT 400 WORDS)