The transmission of substituent effects through a benzene framework has been studied by a novel approach, based on the structural variation of the Ph group in p-Ph-C(6)H(4)-X molecules. The molecular structures of many 4-substituted biphenyls were determined from MO calculations at the HF/6-31G* and B3LYP/6-311++G** levels of theory. The twist angle between the phenyl probe (ring B) and the benzene framework carrying the substituent (ring A) was set at 90° to prevent π-electron transfer from one ring to the other and at 0° to maximize it. The structural variation of the probe is best represented by a linear combination of the internal ring angles, termed S(F)(BIPH(o)) and S(F)(BIPH(c)) for the orthogonal and coplanar conformations of the molecules, respectively. Regression analysis of these parameters using appropriate explanatory variables reveals a composite field effect, a substantial proportion of which is originated by resonance-induced π-charges on the carbon atoms of ring A. Field-induced polarization of the π-system of ring A also contributes to the structural variation of the probe. Thus, the S(F)(BIPH(o)) parameter is very well reproduced by a linear combination of the π-charges on the ortho, meta, and para carbons of ring A, an uncommon example of a quantitative relationship between molecular geometry and electron density distribution. Comparison of S(F)(BIPH(o)) with the gas-phase acidities of para-substituted benzoic acids shows that, while the deprotonating carboxylic probe is more sensitive to π-electron withdrawal than donation, the phenyl probe is equally sensitive to both. While the ability of the orthogonal biphenyl system to exchange π-electrons with the para substituent is the same as that of the benzene ring in Ph-X molecules, an increase by about 18% occurs when the conformation is changed from orthogonal to coplanar. The structural variation of the probe becomes more complicated, however. This is due to π-electron transfer from one ring to the other, which is shown to introduce quadratic terms in the regressions.