load control high-cycle fatigue tests at 427 and 649 exp 0 c were conducted on alloy 718 specimens given various surface finishes. the standard surface preparation for fatigue specimens involves a low-stress grind to minimize the residual surface stresses. a low-stress grind surface was used for generating baseline data; various other surfaces that could be considered feasible for large components fabricated in commercial shops were produced on test specimens, and the high-cycle fatigue strength of each was compared surface finishes produced by belt sanding, grit blasting, fine machining, and electropolishing were examined. surface roughness measurements were taken on typical specimens with each surface finish, and residual stress profiles were measured on three of the surface types. results show little or no difference in fatigue life for the various surfaces and indicate that residual stress profile and grain size are more important factors than surface roughness in determining high-cycle fatigue strength. 12 figures, 5 tables. (era citation 81:025903)