¹⁵N natural abundances of soil total N, roots and mycorrhizas were studied in surface soil profiles in coniferous and broadleaved forests along a transect from central to northern Europe. Under conditions of N limitation in Sweden, there was an increase in δ¹⁵N of soil total N of up to 9% from the uppermost horizon of the organic mor layer down to the upper 0-5 cm of the mineral soil. The δ¹⁵N of roots was only slightly lower than that of soil total N in the upper organic horizon, but further down roots were up to 5% depleted under such conditions. In experimentally N-enriched forest in Sweden, i.e. in plots which have received an average of c. 100 kg N ha^-1 year^-1 for 20 years and which retain less than 50% of this added N in the stand and the soil down to 20 cm depth, and in some forests in central Europe, the increase in δ¹⁵N with depth in soil total N was smaller. An increase in δ¹⁵N of the surface soil was even observed on experimentally N-enriched plots, although other data suggest that the N fertilizer added was depleted in¹⁵N. In such cases roots could be enriched in¹⁵N relative to soil total N, suggesting that labelling of the surface soil is via the pathway: - available pools of N-plant N-litter N. Under N-limiting conditions roots of different species sampled from the same soil horizon showed similar δ¹⁵N. By contrast, in experimentally N-enriched forest δ¹⁵N of roots increased in the sequence: ericaceous dwarf shrubs<Scots pine<grass, suggesting increasing use of inorganic N along the sequence. Complementary studies at the major transect sites had shown that 90-99% of fine tree roots had ectomycorrhizas (ECMs). ECMs were 2% more enriched than corresponding non-mycorrhizal fine roots. Fungal sheaths stripped off ECMs were 2.4-6.4 enriched relative to the remaining root core. It is suggested that a flux of N through ECMs to aboveground parts in N-limited forests would leave ¹⁵N enriched compounds in fungal material, which could contribute to explain the observed δ¹⁵N profiles if fungal material is enriched, because it is a precursor of stable organic matter and recalcitrant N. This could act in addition to the previous explanation of the isotopically lighter soil surface in forests: plant uptake of ¹⁵N-depleted N and its redeposition onto the soil surface by litter-fall.