The implications of alternative disturbance specifications in supply functions have been considered in recent theoretical work but little, if any, empirical evidence has been provided on such specifications. Any empirical investigation of the nature of a supply function disturbance is hampered by considerable uncertainty about the deterministic part of the function. Nevertheless, some progress might be made by considering alternative disturbance specifications in conjunction with a variety of deterministic specifications. Applying this approach to the supply of wheat in southern N.S.W., we find some empirical support for an additive disturbance and, surprisingly, for the use of price, rather than revenue, as an argument In supply functions.