In gravimetric methods for a determination of the Moho geometry, the constant value of the Moho density contract is often adopted. Results of gravimetric and seismic studies, however, showed that the Moho density contrast varies significantly. The assumption of a uniform density contrast thus might yield large errors in the estimated Moho depths. In this study we investigate these errors by comparing the Moho depths determined globally for the uniform and variable models of the Moho density contrast. These two gravimetric results are obtained based on solving the Vening Meinesz-Moritz’s inverse problem of isostasy. The uniform model of the Moho density contrast is defined individually for the continental and oceanic lithosphere to better reproduce the reality. The global data of the lower crust and upper mantle retrieved from the CRUST1.0 seismic crustal model are used to define the variable Moho density contrast. This seismic model is also used to validate both gravimetric solutions. Results of our numerical experiment reveals that the consideration of the variable Moho density contrast improves the agreement between the gravimetric and seismic Moho models; the RMS of differences is 5.4 km (for the uniform density contrast) and 4.7 km (for the variable density contrast).