In animal models it is often possible to block Th1-mediated autoimmunity by
deviating the response to Th2. In this article, Graham Rook and colleagues
argue that this might prove a dangerous strategy in humans, and that we sh
ould distinguish between potentially pathogenic Th2 effector cells, and Th2
-like physiological regulators of Th1 responses. The regulators might be tr
iggered by appropriate Th1-inducing bacterial vaccines.