The term " atopic dermatitis " was given by Wise and Sulzberger (1933) to those inflammatory dermatoses which are intimately associated with other allergic manifestations with a strong hereditary influence, such as hay-fever and asthma, described by Coca and Cooke (1923) as atopic. Patients suffering" from atopic conditions have a high incidence of circulating antibodies of a particular type. These antibodies, known as reagins, are responsible for positive immediate-type hypersensitivity reactions when the patient is skin-tested with a wide range of naturally occurring antigens. These antibodies cannot be demonstrated by in vitro techniques. However, their presence can be shown by the passive transfer technique of Prausnitz and K?stner (1921). They also differ from other circulating antibodies by having a high affinity for skin and being heat-labile. In June 1964 a committee of the World Health Organization defined proteins of animal origin with known antibody activity as immunoglobulins. The human immunoglobulins have been divided into three major groups. Two groups of molecular weight -^150,000 were called IgG (previously y, 7Sy, yss, or y2) and IgA (previously ?2A or YiO The third group consists of the macroglobulins of molecular weight r?900,000 and is called IgM (previously Yi*_ P^m, or 19Sy). Most circulating antibodies occur in the IgG fraction, though the IgM fraction contains red-cell iso-antibodies, saline anti-D (rhesus) antibodies, and immune antibodies directed against the somatic antigens of Salmonella spp. Until recently many workers considered reagin-type anti bodies to be immunoglobulins of the type IgA. However, more recently there has been evidence to suggest that the concept has been somewhat oversimplified, and in fact molecules of the class of IgG might be involved. As there is no doubt that reagin type antibodies are readily demonstrated in patients with atopic eczema, the present study was undertaken to determine the levels of immunoglobulins in patients with atopic eczema as compared with those with other skin conditions, including other patterns of eczema. Until recently this has not been possible because of the lack of a simple quantitative technique for deter mining accurately the levels of these different forms of immunoglobulins (Fahey and McKelvey, 1965). Previous work has been of a semi-quan?i?ative type either by the use of scanning of electrophoresis strips (Tickner, McCabe, and Mier, 1959), or, more recently, by the arbitrary assessment of the degree of precipitation in immunoelectrophoresis (K?lder, 1965).