We have subdivided molecules into the following groups. Small molecules (up to 20 electrons in polyatomic or 50 in diatomic), medium molecules (up to 100 electrons), and large molecules. For small molecules one would expect not only Hartree‐Fock functions but correlated (up to 80%‐90%) functions. For the medium size molecules one can obtain at best Hartree‐Fock functions for the next 2 to 4 years. The correlation correction can be obtained with techniques which relate empirically the electronic density to correlation energy. A simple empirical relation for atomic cases is presented. The medium size molecules will offer the appropriate examples for developing semi‐empirical schemes of extreme simplicity where the knowledge of the molecular structure should in time suffice to give atomic gross charges and orbital energies. For large molecules these are likely the only computational possibility for some time to come, unless the molecule possesses trans‐lational symmetry. The relation between atomic gross charges and the inner shell orbital energies is discussed for the cases of adenine, cytosine, guanine, and thiamine. Copyright © 1969 John Wiley & Sons, Inc.