The literature describing effects of morphine on cells of the immune system points to the clear conclusion that morphine given in vivo suppresses a variety of immune responses that involve the major cell types in the immune system, including natural killer (NK) cells, T cells, B cells, macrophages and polymorphonuclear leukocytes (PMNs). Depression of NK cell activity has been reported in humans, monkeys and rodents. Similarly, responses of T cells are depressed by morphine, as assessed by inhibition of induction of delayed-type hypersensitivity reactions and cytotoxic T-cell activity, modulation of T-cell antigen expression, and depression of responses to T-cell mitogens. Effects on T cells have been reported in humans, monkeys and rodents. Effects of morphine on B-cell activity have mainly been tested in rodents using assays of antibody formation, which also require macrophages and T cells, preventing a conclusion as to the cell type being affected. Consistent effects on phagocytic cell function have been reported in rodents given morphine. In contrast, studies on immunomodulatory effects of morphine added to cells of the immune system in vitro have shown robust effects on some of these cell types, but not others. There is a rich literature demonstrating downregulation of phagocytic cell function by morphine, particularly for human peripheral blood mononuclear cells (PBMCs) and PMNs. Phagocytosis, chemotactic responses, interleukin production, and generation of activated oxygen intermediates and arachidonic acid products have all been reported to be inhibited. On the contrary, the literature does not support direct effects of morphine on NK cell function, is inconclusive concerning effects on B cells, and provides limited evidence for effects on T cells. The divergence between the in vivo and in vitro data suggests that effects on some cells in the immune system observed after in vivo morphine are probably not direct, but mediated. In aggregate, the literature supports the existence of an in vivo neural-immune circuit through which morphine acts to depress the function of all cells of the immune system. Further, there is strong evidence that morphine can directly depress the function of macrophages and PMNs, and modulate expression of one type of T-cell surface marker. There is, however, little evidence for direct effects of morphine on NK cells and B cells. A further complication emerges from reports of immunopotentiation of immune function in in vitro assays using endogenous opioids. The possibility of different receptors for endogenous and exogenous opioids or of interactions among the activated opioid receptors may account for these opposing effects.