The absolute frequencies of B cells-producing A5A idiotype have been determined in vitro by limiting dilution analysis in a culture system in which every LPS-reactive B cell grows into a clone of IgM-secreting cells. Spleen cells from normal A/J mice contain 1 A5A-idiotype-producing B-cell precursor in 2.5 X 10(3) LPS-reactive B cells. Approximately a 10-20-fold increase in frequencies of precursor cells from antigen priming with Strep A-CHO (1 in 2.8 X 10(2)) or from sensitization with IgG1 anti-A5A idiotype (1 in 1.3 X 10(2)). Injection of IgG2 anti-A5A idiotype which has been shown to suppress A5A idiotype in vivo results in only a marginal and maybe insignificant decrease in precursor frequencies (1 in 6.7 X 10(3)). On the other hand, priming does not result in a detectable qualitative difference in the specific precursor cells, since each clone of B cells secretes 30 ng of A5A-bearing Ig within 8 days of culture, regardless of being unprimed or primed. Nearly half of all A5A idiotype-producing clones, both from unprimed as well as from primed mice, show antigen specificity in binding A-CHO. Priming by antigen, therefore, also results in a 10-fold increase in the frequency of idiotype positive B cells without antigen specificity. This result is a prediction of the network hypothesis.