A review is given of the development of experimental evidence for hyperconjugation in the form of electron-release to an unsaturated or electron-deficient centre from saturated H-C and CC bonds. This type of electron-release can contribute regioselectively to the reactivity of organic molecules, and need not involve breakage of the H-C or CC bond; under these circumstances, H-C and CC bonds make similar contributions to the observed effects of alkyl and derived groups, the relative contributions varying with reaction and solvent. In the limit, however, this form of electron-release can result in breakage of the hyperconjugating bond, and in such circumstances (as in transition states leading to proton-loss or 1,2-elimination), H-C bonds become much more effective than CC bonds because of the greater solvation of incipient protons than of incipient alkyl cations. Evidence for the corresponding hyperconjugation by H-O bonds comes especially, but not exclusively, from the relative rates of electrophilic aromatic replacements affected by H-0 and R-O groups. Recent experiments on the effects of these groups on nucleophilic replacements in aromatic side-chains show that the electron-releasing power of the H-O group is quite variable, and suggest that the impact of H-O hyperconjugation is more dependent on the reaction and the solvent than that of H-C hyperconjugation has been shown to be. Reaction pathways in the hydrolysis of R0.C6H2Br2.CHBr2 and RO.C6H3Br.CHBr2, which can lead to products both of replacement and of elimination, are discussed, and isotope effects in H-C and H-O hyperconjugation are compared.