To compare the orientation and development of jointing with the orientation and magnitude of finite strain recorded in the Upper Devonian rocks of the Appalachian plateau, New York, we mapped systematic joint sets on an area of 20,000 km2. In this area, Wedel <1932> mapped folds with limb dips of less than a degree and axes that change strike by 30¿ from 090¿ in the east to 060¿ in the west. We observed two different cross-strike joint sets that maintain their approximate cross-strike position from east to west. Yet, in detail the angle between the sets is 18¿ ¿2¿ in the east and 30¿ ¿4¿ in the west. In many outcrops one joint set parallels the direction of maximum compressive strain (&egr;f) as recorded by deformed fossils, whereas the other joint set never parallels &egr;f. Rare calcite-filled joints are oriented parallel to the direction of &egr;f. In addition, the calcite-filled joints both cut and are cut by solution cleavage. These observations suggest that the joint set paralleling &egr;f formed during the deformation event represented by the deformed fossils. The joint set that does not parallel &egr;f somehow reflects a deformational event other than that producing fossil distortion, as suggested by strain relaxation experiments. In an outcrop of the Machias formation, where the direction of &egr;f is 15¿ from the trace of a cross-strike joint set, the tensor average from 14 subsurface strain relaxation tests shows in situ maximum comprehensive strain (&egr;oc=10-4) parallel to cross-strike joints but not the direction of &egr;f. Strain relaxation is coaxial with a fabric anisotropy indicated by sonic velocity tests. Our ideas is that the orientations of cross-strike joints parallel to &egr;oc were controlled by the same rock property that causes strain relaxation on overcoring, whereas the orientations of the cross-strike joints parallel to &egr;f were controlled by the stress field causing the fossil distortion. Placing these structural data in a regional context allowed us to construct a dynamic and kinematic model for the structural evoution of the New York plateau. The model indicates variation in both boundry conditions and material behavior through a series of four distinct deformational events, which begin prior to lithification and end in the Recent. Thus our analysis suggests that the structural features we have used represent a set of highly sensitive tools for investigating the deformational history of the Appalachian foreland.