Fluorescein (FLU)-specific murine splenic B lymphocytes from nonimmunized adult mice were prepared by the hapten-gelatin fractionation technique and cultured singly or in very small numbers in 10-microliters culture wells. Growth and differentiation to antibody-secreting status were promoted by polymeric FLU-conjugated antigens with or without added T-lymphocyte-derived conditioned media or purified cytokines. In some cultures, 3T3 fibroblasts or CBA/N thymocytes provided a source of filler cells. Anti-FLU antibody formation was detected by a sensitive enzyme-linked immunosorbent assay (ELISA). With an optimal number (around 300) of 3T3 cells per well, up to 77% of the B cells could be induced to produce detectable antibody. The ELISA permitted detection of antibody formation in essentially all wells where B-cell proliferation occurred, and it was more efficient in detecting antibody-forming clones than the hemolytic plaque assay, whether filler cells were present or not. When 10 B cells rather than 1 were included per well, the ELISA, detecting absorbance in standard fashion, provided a useful method for assessment of B-cell growth- and differentiation-promoting factors (BGDF). It was found that 3T3 cells gave less background stimulation than thymus cells, permitting the detection of as little as 1/100th as much BGDF as with thymocytes, thus offering a dynamic range of up to 30 between control absorbance in the absence of factors and the optimal factor level. Use of 3T3 cells also avoids a potential lymphokine cascade. The system has confirmed that interleukin-2 acts as a BGDF, but it has failed to establish an effect of interferon-gamma on B cells. It has also shown the inactivity of a variety of hemopoietic growth factors on B lymphocytes. This system thus promises to be a useful tool in the further analysis of B-lymphocyte activation.