The electron density of CO has been calculated and compared with near Hartree-Fock limit basis sets (11s9p4d) by the Hartree-Fock (HF), second-order Moller-Plesset perturbation theory (MP2), quadratic configuration interaction (QCISD), and several density functional theory (DFT) methods including the local spin density functional approximation (LSDA) and gradient-corrected methods which use the Becke-Perdew (BP) and Perdew-Wang-Perdew (PWP) gradient corrections to the exchange and correlation (XC) terms. Special attention is paid to the topological properties of the electron density and the effect of the auxiliary basis set on the DFT electron density. The electron densities derived from the HF, MP2, and QCISD methods are more localized near the nuclei relative to the LSDA results. The gradient-corrected DFT calculations including Becke-Perdew and Perdew-Wang-Perdew functionals correct the LSDA electron density toward the QCISD results, but the BP and PWP electron densities are observed to have more electron density near the nuclei than the corresponding QCISD electron density with a large basis set (11s9p4d). The Laplacians of the BP and PWP electron densities obtained with the 11s9p4d basis set are very similar to the MP2 and QCISD densities calculated with the same basis set. The Laplacian of the LSDA electron density is, however, very similar to the HF one. The quality of the auxiliary basis set is important for determining the accuracy of the electron density obtained from the density-fitted DFT calculations.