Predicting the potential of soils to supply N is of considerable importance to maximize agricultural N use efficiency and to minimize environmental losses. This research examines and evaluates the current soil testing approach, which uses extractable organic N (EON) fractions to predict soil N supply, using isotopic 15N tracing, multivariate statistics and meta-analytical techniques. Almost all 20 EON fractions that have been developed during recent decades significantly reflect the potential of soils to supply N, in spite of the strong differences in size and composition of EON due to extraction methodology. The EON fractions have therefore been considered as highly bio-available N pools in soil. However, most of them performed either worse than or similarly to total N as predictor of soil N supply, and the uncertainty of the predicted soil N supply (even under controlled environmental conditions) is still too big for serious improvement of fertilizer management. A micro-diffusion method is developed to estimate gross EON fluxes in order to investigate the biochemical basis for observed relationships between EON and soil N supply. The fate of EON fractions in N mineralization, in particular those fractions that are obtained with weak hydrolyzing salt solutions, is comparable to that of dissolved organic N (DON). Both DON and EON can be considered as (intermediate) decomposition waste products in an abiotic and biotic controlled equilibrium with total N. Therefore, their relationship with soil N supply likely reflect that both DON, EON, and soil N supply are mutually dependent on total N. The dependency of soil N supply on methodological and environmental issues strongly encourages more effort to be put into validation and up-scaling, particularly regarding the quantification of the differences between laboratory and field experiments. A combination of soil testing with simulation modeling is necessary to account for the numerous environmental factors controlling so