Previously reported calculations concerned with spin-orbit enhancement of molecularS-Ttransitions have utilized an approximate one-electron spin-orbit Hamiltonian. The approximations are carried over from atomic spin-orbit theory and are justified on the basis of physical reasoning strictly applicable only to diagonal matrix elements. In order to test critically these assumptions, a treatment of spin-orbit enhancement of the formaldehyde3A2↔1A1transition is given using the complete spin-orbit Hamiltonian. A one-electron approximation to the spin-orbit Hamiltonian is found to be conceptually valid. The one-electron approximation which most closely approximates the complete spin-orbit Hamiltonian is the bare nuclear field Hamiltonian. Calculations on the lowest triplet state of the azines yield similar results.