Experiments were carried out to investigate relationships between a number of nitrogen availability indices and biomass nitrogen, in a wide range of Scottish soils, and to establish the source of the nitrogen released. In one experiment, three soils were incubated for a week with high enrichment (99.2 atom %) (15NH4)2SO4 to label the soil biomass. A number of techniques were used to extract N from the labelled soils: extraction with 1 M KCl (Available Mineral-N; MIN-N); fumigation-incubation (Microbial Biomass-N; BIO-N); drying at 70-degrees-C, rewetting, incubating and then measuring MIN-N released (DRY-N); anaerobic incubation, followed by measurement of NH4+ (Anaerobic-N; AN-N); refluxing soils with 2 M KCl for 4 h and measuring the NH4+ released (Hydrolysable-N; CHEM-N); and measuring labelled N uptake by ryegrass (Plant N Uptake; RYE-N). The pool size and isotopic enrichment of the N released by each of the above methods was determined. In 1992, the N contained in 17 soils from a wide range of sites was extracted using the above techniques but without selective labelling. The amounts of N extracted increased in the order of MIN-N < CHEM-N < AN-N < BIO-N < DRY-N much less than total Kjeldahl N. Biomass-N was found to be well correlated with the N extracted by DRY-N (r = 0.82; P < 0.001), AN-N (r = 0.75; P < 0.001) and CHEM-N (r = 0.54; P < 0.01). The results of the N-15 labelling experiment demonstrated that the different techniques resulted in the extraction of different N pools. The correlation between the N-15 enrichment of the N extracted in the microbial biomass and that in the CHEM-N was very low (0.06), whereas the corresponding values for AN-N, DRY-N, MIN-N and RYE-N were 0.91, 0.97, 0.85 and 0.93. The biological extraction techniques seem to have extracted N at least partly from the microbial biomass pool. The chemical extraction technique would appear to have extracted nitrogen from a quite different pool, or pools, much more closely related to the total soil nitrogen pool (r = 0.95; P < 0.001).