A major aim in immunology has been to understand how the immune system evokes characteristic responses to infection, foreign tissue grafts and tumours. The current view of immunoregulation is based mainly on studies of lymphocyte subsets, either in vitro^1–3 or by adoptive transfer to irradiated recipients⁴. Many reagents are available for defining T-cell subsets^5–8, but only recently have there been helper T-cell-specific antibodies^9,10 against the mouse equivalent of the Leu3/T4 (man) and W3/25 (rat) antigens. It is clear that monoclonal antibodies¹¹ will eventually replace antilymphocyte globulin for immunosuppression in organ grafting^12,13, but although there has been some clinical success^14–16, most monoclonal reagents cause only transient reductions in their target cells in vivo^17–20. This uncertainty in the potency of monoclonal antibodies has led some workers to consider them as targeting agents for such highly cytotoxic drugs as ricin A (ref. 21). We show here that unmodified monoclonal antibodies can be extremely effective at depleting cells in vivo and can be used for the selective manipulation of different aspects of the immune response.