In order to improve the interpolation accuracy of the absolute linear encoder, a compensating method for the interpolation error is presented. By using the Fourier theory to analyze the A and B moire fringe signals, the moire fringe signal models for the absolute linear encoder is set up. By correcting the phase, amplitude, harmonics and direct current components of the signal modes, the ideal signal models are obtained. By comparing the interpolation position of the practical and ideal signal models, the interpolation error of the absolute linear encoder is obtained. After that, by compensating the interpolation position according to the obtained error, the interpolation accuracy can be improved remarkably. The JC09 absolute linear encoder is used to test and verify the presented method. The test results show that the interpolation relative error can be decreased from 2.70% to 1.05% . It demonstrates that this method can improve the interpolation accuracy significantly. In addition, due to the simple theory, the method can be easily realized.