The reliability of various quantum-chemical approaches for the calculation of bulk properties of lithium tetraborate Li2B4O7 was examined. Lattice parameters and the electronic structure obtained with densityfunctional theory (DFT), with DFT-Hartree-Fock (HF) hybrid methods,, and with the semiempirical method MSINDO were compared to available experimental data. We also compared the results at DFT level using different wave functions, either based on linear combinations of atom-centered orbitals (LCAO), or on plane waves, as implemented in the crystalline orbital programs CRYSTAL and VASP. The basis set dependence of calculated properties was investigated for the LCAO method. In the plane wave approach ultrasoft pseudopotentials (US PP), and projector-augmented wave (PAW) potentials were used to represent the core electrons. For all methods under consideration, the calculated Li2B4O7 structure parameters are close to each other and agree within a few percent with measured values. A more pronounced method dependence was found for the band structure, the band gap and the cohesive energy. Closest agreement between theoretical and experimental results for the band gap was obtained with the DFT-14F hybrid methods while pure DFT methods underestimate and HF based methods overestimate the measured value. It was found that the calculated band gap strongly depends on the atomic basis set in the LCAO approach. The description of the core electrons considerably affects the cohesive energy obtained with the plane wave approach. Atomic charges based on a Mulliken analysis were compared to effective charges obtained from Raman spectroscopy. Electron density maps are used to analyze the character of B-O and Li-O interactions.