It is usually assumed that the maximal noncovalent bond strength is achieved by full geometry optimization of the geometry of the dyad. Density functional theory calculations show this not to be the case. A number of systems are considered that include osme, tetrel, pnictogen, and chalcogen bonds, involving both σand π-holes, as well as hypervalency. By suitable adjustment of the bond angles within the Lewis acid, the base can be drawn closer than in the optimized structure, with an accompanying substantial strengthening of the noncovalent bond, by more than 10 kcal/mol in some cases. The energetic cost of this deformation from the optimized geometry can be surprisingly small in comparison to the gain in the interaction energy. Taking the opposite approach of first pushing the two subunits closer to one another and then permitting internal geometries to adjust to the shortened distance produces only minimal bond strengthening.