Several accurate analyses have revealed a statistically significant north-south ecliptic asymmetry in the large-angle correlations strength of the cosmic microwave background (CMB) radiation temperature field data from the Wilkinson Microwave Anisotropy Probe (WMAP). This asymmetry is inconsistent with the statistical isotropy expected in the concordance cosmological model $\ensuremath{\Lambda}\mathrm{CDM}$. It has been suggested that a possible cause-effect relationship exists between this large-angle anisotropy and the anomalous CMB quadrupole-octopole planes alignment. In turn, this later phenomenon (or both) would be a consequence of one or more of the following undesired effects in CMB data: a systematic error in the data processing or in the instrument characterization, residual foregrounds, and large-angle correlations induced by the incomplete sky CMB data (cut-sky masks are needed to reject galactic foregrounds). Here, it is proved that the north-south asymmetry is unrelated to the quadrupole ($\ensuremath{\ell}=2$) and the octopole ($\ensuremath{\ell}=3$) properties because we find, at high confidence levels, such large-angle anisotropy in three- and five-year WMAP CMB maps containing only the multipole components $4\ensuremath{\le}\ensuremath{\ell}\ensuremath{\le}10$. The statistical significance depends on both, the CMB map analyzed as well as the cut-sky mask applied to exclude foregrounds. In general, we obtain that the significance level of the north-south asymmetry is less in five-year WMAP data with KQ75 ($\ensuremath{\gtrsim}90%\text{ }\text{ }\mathrm{CL}$) than it is in three-year data with Kp0 ($\ensuremath{\gtrsim}96%\text{ }\text{ }\mathrm{CL}$). For instance, in the WMAP internal linear combination (ILC)-five-year map with the KQ75 mask (a sky cut of 28.4%) this phenomenon is observed at 92.7% CL, whereas for the WMAP ILC-three-year map with the Kp0 mask (a sky cut of 23.5%) this phenomenon appears at 96.5% CL. Moreover, it is also shown that this hemispherical asymmetry is unlikely due to systematics or foreground contaminants, because it is present in single-frequency, multifrequency, and cleaned ILC-type CMB maps. Additionally, robustness tests show that the statistical significance of the north-south asymmetry is affected by the use of the KQ85 and KQ75 masks in five-year single-frequency WMAP CMB maps, whereas it is insensitive with respect to application of the Kp2 and Kp0 masks in three-year single-frequency WMAP CMB maps. The confidence levels were obtained using sets of Monte Carlo CMB maps produced according to the $\ensuremath{\Lambda}\mathrm{CDM}$ model.