When stimulated by the lymphokine interleukin 2, human intestinal mucosal mononuclear cells mediate lymphokine-activated killer cell activity. When supplied with optimal doses of exogenous interleukin 2, lamina propria mononuclear cells isolated from inflammatory bowel disease and control tissue display comparable levels of cytotoxicity in vitro. However, cultures of Crohn's disease- and ulcerative colitis-derived cells contain significantly decreased interleukin 2 activity, suggesting that in vivo the availability of interleukin 2 may be limited, perhaps resulting in impaired cytotoxic function. To test this hypothesis, lamina propria mononuclear cells from inflammatory bowel disease and control patients were stimulated to produce endogenous interleukin 2, which was then used to induce autologous lymphokine-activated killer cells. When tested against K562 and Daudi target cells, Crohn's disease cells, despite producing only one-third of the amount of interleukin 2 generated by control cells, exhibited comparable levels of cytotoxicity. In contrast, ulcerative colitis cells produced substantially less interleukin 2 and exhibited remarkably low lymphokine-activated killer cell activity. When the same cells were supplied with an amount of human recombinant interleukin 2 equivalent to the average titer found in control cultures, similar results were obtained, and Crohn's disease cells even showed a significantly greater cytolytic activity than controls. These results suggest that the observed differences in lymphokine-activated killer cell activity cannot be attributed to the level of interleukin 2 alone, and that response to this lymphokine is different among Crohn's disease, ulcerative colitis, and control intestine. In Crohn's disease, there is either an increased number of interleukin 2-responsive cells or an exacerbated reactivity to interleukin 2. In ulcerative colitis, a loss of interleukin 2-responsive cells, a hyporesponsiveness to interleukin 2, or both might be present. In conclusion, this study demonstrates that reactivity to interleukin 2 distinguishes inflammatory bowel disease from control intestinal mononuclear cells, and, under appropriate experimental conditions, it can be used to uncover abnormalities of intestinal immunity.