Feedback error learning has an architecture of two-degree-of-freedom control in which the feedforward controller is tuned to be the inverse model of the plant. Recently, Miyamura and Kimura have established a control theoretical validity of the FEL method in the framework of adaptive control and proved its stability based on the strictly positive realness. In this paper, we propose a design method for the two-degree-of-freedom control system, which does not require any positive real conditions. The plant is described by a transfer function which does not have proper inverse.