Mathematical processing associated with solving short geometry problems requiring logical inference was examined among students who differ in their levels of general giftedness (G) and excellence in mathematics (EM) using ERP research methodology. Sixty-seven male adolescents formed four major research groups designed according to various combinations of G and EM factors, while another group of seven students were considered to be "super mathematically gifted" (S-MG). EM and G factors affected behavioral measures similarly: EM and G students were more accurate than their non-EM and non-G counterparts respectively. At the same time, G and EM factors affected electrophysiological measures differently, with significant interaction between G and EM factors associated with absolute ERP amplitudes at some of the solution stages. S-MG students exhibited significantly lower absolute ERP amplitude values, which we attribute to a neural efficiency effect in this research group. Based on the differences revealed, we suggest that our research demonstrates that G and EM factors are interrelated individual traits which are different in nature. Since the accuracy and strength of electrical potentials associated with solving the problems appeared to be of an accumulative nature, we argue that S-MG is an extreme expression of combined EM and G factors. Taking it one step further, we suggest that ability grouping in school mathematics must take into account both EM and G factors.