We have attempted to reproduce the findings of Benveniste and co-workers, who reported in 1988 that degranulation of human basophil leukocytes is triggered by very dilute (10(2)-10(120)) antiserum against IgE. The results were contrary to conventional scientific theory and were not satisfactorily explained. Following as closely as possible the methods of the original study, we can find no evidence for any periodic or polynomial change of degranulation as a function of anti-IgE dilution. Our results contain a source of variation for which we cannot account, but no aspect of the data is consistent with the previously published claims.