We discuss some property-property relationships (PPRs) for the estimation of vapour pressure with regard to their predictive power in terms of accuracy and applicability. Seven different PPRs mostly based on the Clausius-Clapeyron equation are analysed, including for example the method of Mackay or those of Fishtine/Vetere. A data set of 375 compounds was compiled, which contains all required quantities as measured ones. Several criteria are defined to determine the accuracy of the estimation methods. By applying partial order theory, it turns out that two equations by Fishtine/Vetere are optimal, i. e. yield the best estimation results, albeit they are not comparable. According to the main focus of this paper our data set is discussed in terms of chemical structures. Structural elements, such as carbonyl-function, aminefunction, halogen substitution etc., were identified, which were suitable to characterize the diversity of a set of organic chemicals. Given the composition of the current data set, it was necessary to reduce the number of structural elements to eight. Finally, we introduce the admatrix, describing the quality of an estimation equation with respect to a certain structural element. The numerical differences of the adij-values among different estimation equations are not large. Hence a fuzzy partial order approach was applied to get the best vapour pressure estimation equations with respect to a distinct structural element. Four estimation methods can be recommended, concerning their accuracy with respect to different chemical structures.