A method previously developed for predicting the coupled magnetoelectric effect in a piezoelectric–piezomagnetic composite containing ellipsoidal inclusions is a general, explicit method and not as Huang and Kuo [J. Appl. Phys.81, 1378 (1997)] reviewed. After correcting two obvious mistakes in Huang and Kuo’s numerical results for effective coupled magnetoelectric coefficients, their results appear to be the same as one type of previous results. ©1997 American Institute of Physics.