Saddlepoint techniques have been used successfully in many applications because of the high accuracy with which they can approximate intractable densities. In this work through considering the integer-valued bilinear model, the saddlepoint method is applied which provides an efficient approach to parameter estimation. Although the conditional least squares and quasi-maximum likelihood methods could be used to estimate the parameters of the model, the saddlepoint method is adopted as it was still expected to have a better performance than the conditional least squares and quasi- maximum likelihood methods in practice.
