The results of plant breeding trials with populations of fodder pea strains and broad bean hybrids were the basis of consideration on the interrelationship between some traits - the yield structure elements. Developed by Eaton, a relatively new method of yield component analysis called the two-dimensional partitioning method (TDP) was applied to analyse the data. The method, which combines multiple regression and ANOVA, allows for concise tabular presentation and simple interpretation of the distribution of traits in one direction and the sources of variance according to ANOVA model in the other direction. Additionally, the interpretation of the results was supported by such standard statistical techniques as ANOVA, simple and multiple regression and path analysis. The main components of pea yielding were plant height and the number of pods per plant. Among the analysed characters of broad bean the number of nodes with pods on the main stem, which turned out to be the determinant of broad bean yielding, might be strongly affected by environmental conditions. The number of nodes with pods might be considered a selecting character of high potential yielding of broad bean genotypes.