The anti-human serum albumin (HSA) B-cell repertoire of C57BL/6 mice was examined by culturing splenocytes at limiting dilution following polyclonal stimulation with Escherichia coli lipopolysaccharide and a lymphokine mixture. The frequency of anti-HSA precursors was determined before and after immunization with alum-precipitated HSA and 10(9) killed Bordetella pertussis organisms, by submitting clonal supernatants to an ELISA. Anti-HSA IgG1-forming precursors were rare in unimmunized spleens, representing approximately equal to 1 in 500,000 splenocytes or only approximately equal to 100 cells per spleen. Between day 5 and day 7 after immunization, this figure increased to approximately equal to 20,000 cells per spleen. Over the following 3 weeks, there was a progressive increase in the mean optical density generated in the clonal ELISA, presumably due to affinity maturation of the B-cell population. When freshly deaggregated HSA was injected before or even up to 4 days after challenge immunization, the appearance of anti-HSA IgG1-forming cell precursors was largely prevented. The effect was most marked with 5 mg or 1 mg of soluble HSA, but impressive partial effects could be seen with as little as 10 micrograms of HSA if administered before challenge immunization. Most of the few clones seen after the higher doses of the toleragen appeared to make antibody of low affinity. The capacity to influence the B-cell pool by soluble antigen administered just 1-2 days before the sudden appearance of IgG1 precursors argues against the totality of the effect being due to T-cell-mediated suppression and in favor of a direct effect on B cells.