"Neural efficiency" hypothesis posits that neural activity is reduced in experts. Here we tested the hypothesis that compared with non-athletes, elite athletes are characterized by a reduction of cortical activation during an engaging upright standing. EEG (56 channels; Be-plus Eb-Neuro and stabilogram (RGM) data were simultaneously recorded in 10 elite karate, 10 elite fencing athletes, and 12 non-athletes during a simple bipodalic (standard Romberg) and a more engaging monopodalic upright standing. Balance was indexed by body "sway area". The EEG data were spatially enhanced by surface Laplacian estimation. Cortical activity was indexed by task-related power decrease (TRPD) of EEG alpha power (8-12Hz) during monopodalic referenced to bipodalic condition. The body "sway area" was larger during the monopodalic than bipodalic upright standing in all groups. Low-frequency alpha TRPD (about 8-10Hz) was lower in amplitude in the karate and fencing athletes than in the non-athletes at left central, right central, middle parietal, and right parietal areas (p<0.01). Similarly, the amplitude of high-frequency alpha TRPD (10-12Hz) was lower in the karate and fencing athletes than in the non-athletes at right frontal, left central, right central, and middle parietal areas (p<0.03). These results suggest that during monopodalic referenced to less engaging bipodalic condition, the power decrease (i.e. the desynchronization) of cortical activity at alpha rhythms is largely reduced in elite athletes than in non-athletes, in line with the "neural efficiency" hypothesis. The present study extends our understanding of the physiological mechanisms at the basis of the "neural efficiency" for engaging upright standing in elite athletes.