Some time ago Sicher [l] suggested that the ratio of the rates of oxidation of an epimeric pair of alcohols with chromic acid could be regarded as an approximate value for the free energy diffeience AG of these alcohols through the relationshipl)This equation turned out to be of fundamental importance for the understanding of the steric effects operating in the alcohol oxidation. The relationship was verified later by Wilcox [Z] Recently we found a relationship between oxidation rates and the strain difference between alcohols and the corresponding ketones, as evaluated by means of molecular mechanics [5]. It allows to rationalize oxidation rates within a factor of ca. 2 in the average. However, some compounds correlate badly, and the exceptions can only in part be explained by deviations of the entropies of activation from the normal value [6]. From this viewpoint the correlation of Sicher is of particular interest. I t correlates two experimental quantities, so that no assumptions concerning the reaction mechanism have to be made. Furthermore, it only considers the strain difference between epimeric alcohols, leading to one and the same ketone. The strain of the transition states needs not to be known, so that the more difficult part oi the problem is eliminated. One might therefore expect that the success (or failure) of the Sicher correlation should pro\ ide a reasonable estimation of the minimal uncertainties one would encounter in auy other rationalization of oxidation rates such as the one we used ourselves.Unfortunately, the general validity of Sacher's relationship is questionable. I t is based on 10 pairs of epimeric alcohols, 8 of which are cyclohexanols. The k,/k,-ratios of 9 of the epimeric pairs are in the range of 1.7 to 6.6, with a single value at 33.6, and the rate constants are spread out over a range of only 65. Therefore it appeared 1) a and e refer t o the more (a) and less (c) strained alcohol.