T object of this investigation was to develop ''tear-out design criteria by means of which the strength of the material outside the bolt or rivet in a joint in tension could be more accurately determined. As test data were obtained and studied it became apparent that the edge distance had a definite relationship with the bearing strength and the test program accordingly was arranged to show how the bearing strength varies with the edge distance, with bolt or rivet size and with sheet thickness. A number of typical test specimens are shown in Fig. 1. In all cases the specimen was made of two pieces of 24ST aluminum alloy or chrome-molybdenum steel sheet fastened together with a single bolt or rivet. The bolts were of steel heat-treated to an ultimate tensile strength of 125,000 lbs. per sq.in. and were used in all cases in double shear. The rivets were made from 17S aluminum-alloy rivet wire purchased under Army-Navy specifications and heat treated in a nitrate bath to an ultimate tensile strength of 55,000 lbs. per sq.in. They were used in all specimens in single shear. All specimens were tested in tension to failure. Most of the specimens were formed with square ends. A few, however, were made with rounded ends. Of the latter specimens some had the bolt hole drilled concentric with the rounded end, some had the hole drilled with a plus offset and some with a minus offset, noted in the presentation of the test results. To clarify the offset designations it may be stated that when the distance from the center of the hole to the end of the lug was less than the radius of curvature of the rounded end, the offset was assumed to be minus. The sheet stock used in these tests was purchased under Army-Navy specifications. Test coupons were cut from the 24ST material and showed an ultimate tensile strength that varied from 68,100 to 73,700 and averaged 71,000 lbs. per sq.in. The bearing values obtained in the tests were not corrected to a common ultimate strength. In the case of the chrome-molybdenum steel specimens tested in the as received condition, the test coupons showed a variation in ultimate strength of 71,800 to 83,500 lbs. per sq.in. The