SUMMARY A radioimmunoassay for antibody to Candida albicans is described. The test uses whole, killed organisms as the antigen and radiolabelled sheep anti-human globulins to quantitate different classes of antibody to C. albicans. The assay has been compared with an Ouchterlony precipitin method and found to be simpler, more rapid, and more sensitive than the latter. Results obtained from two groups of symptomless adults indicated that the range of antibody level was wider for a hospitalised group than for a group of blood transfusion donors, particularly in respect of IgG and IgA antibody. The reason for the increase of antibody in hospital patients was not clear but may have been related to antibiotic therapy. The difficulties in interpretation of Candida serology have therefore been re-assessed in the light of more detailed knowledge of the range and type of antibody to be expected in normal individuals.Candida albicans infection presents many problems in clinical practice. Most notable of these is the distinction between serious organ infection involving, for example, the lung or oesophagus and simple colonisation of these sites. This distinction is difficult to make by cultural examination, although the demonstration of mycelial forms in the specimen or biopsy material has been regarded as an important indication of infection (Winner, 1969). Recently, it has been suggested that the demonstration of serum antibody to mycelial rather than blastospore antigens, using a fluorescent antibody method, is a simpler and more sensitive way of detecting deep tissue infection (Ho et al., 1976). Heavy colonisation of the mouth, pharynx, gut, and urinary tract by the organism is now commonly seen in patients receiving broad-spectrum antibiotics (Seelig, 1966)