The assessment of sperm immunity must take into account the primary role of antibodies in local genital tract fluids rather than in serum, and the observation that antibody level may fluctuate markedly with time. Therapy for sperm antibodies in the female partner is complicated by their multiple potential sites and modes of action, and by our incomplete understanding of their clinical significance. Thus, if present in serum as well as in cervical mucus, their effects are unlikely to be overcome by intrauterine insemination, since the serum antibodies are likely to be represented in tubal fluid by transudation. Moreover, even when they are manifest solely at a cervical level (in mucus), there is little evidence of effectiveness either for this method of treatment or for others, including occlusion (condom) therapy to 'remove the immunogenic stimulus', antibiotics to eliminate any adjuvant effects of cervical infection and mild immuno-suppression.