The "meta effect" was defined decades ago as transmission of electron density between meta situated groups on benzenoid compounds in the singlet excited state. This contrasts with the well-accepted ortho-para transmission in the ground state. Our early theoretical efforts made use of the relatively primitive computational methodology available at the time but were supported by our experimental studies. The operation of the "meta effect" has been questioned, and the purpose of the present investigation was to make use of the more sophisticated theoretical methods now available. Thus, ab initio computations of m- and p-methoxybenzyl cations and radicals were carried out at the CASSCF(8,8)/G-31G* level. Similar computations on acetate anion and acetoxy radicals were obtained in order to provide energies of the methoxybenzyl ion pairs and also the radical pairs. This information bears on the initial competition between S-1 heterolysis and homolysis as well as preferences for meta versus para scission. The evidence provided shows that the energetic preference is for heterolysis to give ion pairs rather than radical pairs, and that the S-1 m-melhoxybenzyl cation is considerably lower in energy than the S-1 p-methoxybenzyl cation, thus confirming the earlier more primitive computations. Further, the 3,5-dimethoxybenzyl cation is even more heavily stabilized.