For many years the antistreptolysin O (ASO) z titer has been the most commonly used immunologic indicator of infection with group A streptococci. Within the last decade certain limitations in using the ASO response to define previous streptococcal infection have been recognized. The failure to demonstrate an ASO response in about 20% of patients with streptococcal infection of the upper respiratory tract has stimulated interest in antibodies to other streptococcal antigens (1, 2). Perhaps more importantly, and certainly more provocatively, reports from this and other laboratories suggested that ASO titers were often not elevated or only slightly elevated in children with streptococcal pyoderma when compared with children in the same or a similar population (3-5). A further study from this laboratory in which children were followed carefully with serial cultures and bleedings showed that in contrast to the antistreptococcal deoxyribonuclease B (anti-DNase B) response, children with streptococcal pyoderma generally exhibited a feeble ASO response (6). This study has been subsequently confirmed by others (7,8).Initially, several explanations for this observation seemed possible. A generalized immunological unresponsiveness to all streptococcal extracellular antigens was ruled out by the usually excellent anti-DNase B response in patients with streptococcal pyoderma (6-8). The possibility that the poor ASO response was due to low production of this antigen by the pyoderma strains seemed unlikely since no difference in in vitro production was demonstrated between strains recovered from the respiratory tract and those infecting the skin (6). With the failure to find support for these two possibilities, a third hypothesis was considered: that properties of the skin per se might be responsible for this peculiar disparity in the ASO response.It has been known for some time that cholesterol is a potent inhibitor of two of