We have reexamined the ability of anti-human IgG antibodies to induce histamine release from human basophils. A panel of purified murine mAbs with International Union of Immunological Societies-documented specificity for each of the four subclasses of human IgG was used. Of the 24 allergic subjects studied, the basophils of 75% (18/24) released greater than 10% histamine to one or more anti-IgG1-4 mAb, whereas none of the 13 nonatopic donor's basophils released histamine after stimulation with optimal amounts of anti-IgG mAb. The basophils of 85% (11/13) of the nonatopic donors did respond to anti-IgE challenge, as did 92% (22/24) of the atopic donor cells. Histamine release was induced most frequently by anti-IgG3, and 10/18 anti-IgG responder cells released histamine with mAb specific for two or more different subclass specificities. The rank order for induction of histamine release was anti-IgG3 greater than anti-IgG2 greater than IgG1 greater than anti-IgG4. As in our previous study using polyclonal anti-IgG, 100- to 300-micrograms/ml quantities of the anti-IgG mAb were required for maximal histamine release, about 1000-fold higher than those for comparable release with anti-human IgE. Specificity studies using both immunoassays and inhibition studies with IgE myeloma protein indicated that anti-IgG induced histamine release was not caused by cross-reactivity with IgE. Ig receptors were opened by lactic acid treatment so that the cells could be passively sensitized. Neither IgE myeloma nor IgG myeloma (up to 15 mg/ml) proteins could restore the response to anti-IgG mAb. However, sera from individuals with leukocytes that released histamine upon challenge with anti-IgG mAb could passively sensitize acid-treated leukocytes from both anti-IgG responder and nonresponder donors for an anti-IgG response. The only anti-IgG mAb that induced release from these passively sensitized cells were those to which the serum donor was responsive. Sera from non-IgG responders could not restore an anti-IgG response. These data led to the hypothesis that the IgG specific mAb were binding to IgG-IgE complexes that were attached to the basophil through IgE bound to the IgE receptor. This was shown to be correct because passive sensitization to anti-IgG could be blocked by previous exposure of the basophils to IgE. We conclude that anti-IgG-induced release occurs as a result of binding to IgG anti-IgE antibodies and cross-linking of the IgE receptors on basophils.