In accordance with recent literature data, it is proposed that an increment of + 2.3 cps and + 1.8 cps is to be added to the vicinal PJ) and gem (2J) coupling values respectively, each time an a-oxygen (or a-nitrogen) atom has one of its free electron pair p-orbitals parallel with the carbon hydrogen bond of one of the protons involved in the coupling under consideration. An influence on 3J coupling values in three membered rings (epoxides) is however not to be considered. Molecular deformations which disturb the parallelity (i.e. in the rigid half chair form of pentacyclic compounds or in the twist form of hexacyclic compounds) nullify this effect. The Karplus rule and the electronegativity rule have regained a great amount of their reliability with this supplementary effect.The generalised deduction that electronegative substitution causes a marked algebraic decrease of the vicinal and geminal coupling constants ( I ) has been controversed recently (').There is indeed an increasing amount of data which is in contradiction with these electronegativity rules. On the other hand it is also a fact that the so valuable "Karplus" effect (3) is often incapable to explain observed coupling values. This creates difficulties in evaluating molecular shapes and conformational data. Unfortunately the deviations seem to be greatest in cyclic compounds with more or less rigid structure, which are of the most fundamental character, (l) J.R.