The regulatory biology of antibody formation entered a new phase of study with the development of selective theories of immunity. The discovery of the 'one cell - one antibody' dogma and the demonstration that only a small minority of B cells possessed receptors specific for a given antigen were consistent with Burnet's clonal selection hypothesis, which was later formally proven by preparing antigen-specific lymphocytes and inducing clonal activation in vitro. Clonal analysis has aided precise study of immunoregulation for both B and T lymphocytes. Clonal activation of B cells in the absence of T cells is now possible with high cloning efficiency. It requires the combined action of certain antigens and growth factors, collectively termed B-cell stimulatory factors (BSFS). Single cell analysis has shown that most BSFS so far tested, in contrast to most claims in the literature, possess the capacity (in synergy with antigen) to: stimulate B cells out of the G0 phase into active cell cycle; promote sequential mitotic divisions; and induce differentiation to active secretory status. This is clearly true for IL-1, IL-2, and BSF-p2. These multiple actions resemble those of the colony-stimulating factors in haemopoiesis. Regulation of antibody production by T lymphocytes can also be profitably analysed in clonal systems. The immunoregulatory problem of tolerance can also be analysed by means of clonal techniques. Studies are summarized which indicate that T-cell-mediated suppression and functional silencing of toleragen-specific lymphocytes are both cooperatively involved in many tolerance models. For the B lymphocyte, tolerance can be induced without an actual deletion of the cell involved; rather, the tolerant cell appears to have received and stored a negative signal, rendering it unresponsive to normally immunogenic stimuli. Thus, a state termed 'clonal anergy' has been induced within the cell. Functional clonal deletion has also been noted in several models to T-lymphocyte tolerance, but here it is not known whether clonal anergy or actual death of the relevant cell is at work. Self-tolerance sufficient to be consistent with good health need not mean a total absence of cells with any degree of self-reactivity. Indeed, it is clear that some B cells capable of forming antibody with some degree of affinity for self-constituents exist in the body, and can be activated, for example by lipopolysaccharide. The requirement is to limit the amount, affinity and duration of autoantibody production. A model suggesting how this may be achieved is presented.