In a previous study, alimentary secretions and sera were obtained from 49 human subjects (1). Immunoglobulin A was found to predominate in parotid saliva. Duodenal specimens were reported to contain IgA, IgM, IgG and IgD. The apparent lack of a uniform distribution of the four major classes of immunoglobulins led us to feel that this was a real phenomenon. This seemed to be supported by the association of locally produced anti-BSA with immunoglobulins other than IgG in 9 of 31 subjects without circulating anti-BSA, as well as apparent antibody associated with IgG, IgA and IgM in duodenal contents from most of the subjects with high amounts of circulating anti-BSA. The purpose of this communication is to present follow-up data which demonstrate that the conclusions reached in the previous report were based on what now appear to be false-positive immunodiffusion and radio-immunodiffusion studies.