The first part of this article discusses the isolation and characterization of several monoclonal antibodies to the B lineage-specific surface molecule, B220. B220 was shown to be expressed on precursors of B cells by the demonstration that removal of B220+ cells from B cell-depleted bone marrow removes the ability of bone marrow to regenerate B cells. Although these antibodies recognize a broad range of differentiation stages within the B lineage, they can be used to isolate highly enriched populations of pre-B cells from mouse bone marrow. We also describe the use of antibodies to the surface markers B220 and ThB to define two sequential stages of pre-B cell differentiation. A simplified diagram of our current view of the B lineage differentiation sequence is shown in Figure 4. No attempt has been made to include the various functionally defined B cell subsets on this diagram since we know almost nothing about the expression of these two surface markers on them. This model reflects an assumption that the early part of B cell differentiation is a linear rather than a branching pathway. At present, there is no evidence for a branching pathway, but little evidence against it either. B220 is the first B cell-specific molecule known to be expressed during differentiation and it continues to be expressed on most subsequent B lineage cells. In this regard, it resembles the Thy-1 molecule on thymus-derived lymphocytes and, like Thy-1, B220 should be quite useful for identifying and classifying B lineage cells. One example of this is the use of B220 expression to clearly assign germinal center cells to the B lineage. The example of Thy-1+, RA3-2C2+ cells from mice with the lpr/lpr genotype, however, suggests that some caution should be used when interpreting data, especially with pathological samples. The availability of substantially purified pre-B cell populations has made it possible to follow changes in immunoglobulin gene organization and expression during differentiation. Our current understanding of these events is also shown in Figure 4, correlated with cell surface phenotype. The large pre-B cell population has extensive heavy chain rearrangements and synthesizes significant quantities of mu heavy chain, but does not yet have detectable light chain gene rearrangement. The small pre-B population consists of two cell types, some with kappa gene rearrangement and some without. This suggests that kappa rearrangement occurs within this cell population, which is homogeneous with respect to morphology and surface phenotype. The asynchrony of heavy and light chain gene rearrangement results in an asynchrony at the level of expression of these genes as well, but the purpose of this remains one of many unanswered questions about pre-B cell differentiation. Now that it is possible to identify, isolate, and manipulate pre-B cells as readily as B or T cells, many of these questions may now be addressed.