The mechanisms by which endothelium-dependent relaxants and nitrovasodilators cause relaxation of vascular smooth muscle has been reviewed. A model explaining these observations is summarized in Fig. 1. The endothelium-dependent vasodilators through interaction with their appropriate receptors are thought to activate phospholipase A2 and cause the release of an unsaturated fatty acid. The released unsaturated fatty acid or a metabolite is thought to be the "endothelial relaxant factor" that interacts with the smooth muscle component to cause relaxation. While the unsaturated fatty acid may be oxidized in either the endothelial cell or smooth muscle cell, the lability of the endothelial relaxant factor suggests that at least some of this processing occurs before its release from the endothelium. the model in Figure 1 suggests that an oxidized fatty acid or a derived free radical is responsible for activation of smooth muscle guanylate cyclase and increases in cyclic GMP levels. As pointed out above, the use of various inhibitors of fatty acid release and metabolism has not allowed us or others to predict the structure of the active material. To date the best evidence suggests that the unsaturated fatty acid is a product of either the lipoxygenase or P-450 pathways. Nitrovasodilators are thought to form nitric oxide free radical and directly activate guanylate cyclase as described above. Activated guanylate cyclase, whether by endothelium dependent agents or the nitrovasodilators, then increases the formation of cyclic GMP, which activates cyclic GMP-dependent protein kinase. The phosphorylation state of various proteins is then altered and, eventually, myosin light chain is dephosphorylated and relaxation occurs. Whether this mechanism involves cyclic GMP-dependent changes in activities of myosin light chain kinase and/or myosin light chain phosphatase remains to be determined. Although the altered phosphorylation state of myosin light chain that results from cyclic GMP accumulation may explain the mechanisms of action of cyclic GMP in smooth muscle relaxation, other mechanisms can not be excluded. For example, some additional studies which we have not summarized here indicate that the integrity of the membrane and Na+-K+ pump can modify both cyclic GMP synthesis and relaxation in rat aorta (38 and unpublished observations). Apparently complex interactions may exist in smooth muscle and other tissues which regulate cyclic GMP accumulation and/or its expression on some process. While several functions for cyclic GMP have been suggested, there is considerable evidence which suggests that one of its roles is relaxation of airway and vascular smooth muscle.