Birch (Betulapendula) seedlings were grown in pots of unsterile brown earth, vermiculite-peat mixture or coal spoil supplemented with freshly deposited basidiospores of ectomycorrhizal fungi. Species of Hebeloma, Laccaria and Inocybe established mycorrhizas from spores and prevented mycorrhiza formation by other fungi occurring naturally in (or possibly contaminating) the rooting media. Species of Lactarius, Cortinarius, Leccinum and Russula, and Amanita muscaria, Elaphomyces muricatus, Scleroderma citrinum and Suillus luteus did not establish mycorrhizas from spores. Paxillus involutus formed mycorrhizas from spores in coal spoil but only poorly in brown earth, demonstrating some substrate-dependency in mycorrhizal establishment. However, spores of Lactarius turpis and Leccinum scabrum from fruit bodies on coal spoil did not establish mycorrhizas on birch seedlings in coal spoil, whereas those of P. involutus did so. Also, seedlings planted in intact coal spoil samples from beneath fruit bodies of Amanita muscaria, Leccinum roseofractum and Scleroderma citrinum did not develop mycorrhizas of these fungi, whereas samples from beneath fruit bodies of Inocybe lacera and P. involutus did so. Thus, soil-effects did not obscure a major difference between two groups of mycorrhizal fungi, i.e. those that could or could not establish mycorrhizas on seedlings from spores added to unsterile soil. At least for the fungi tested, these seem to be generic attributes, and they correlate with groups termed early- and late-stage in successions of mycorrhizas. In limited tests, Hebeloma crustuliniforme established mycorrhizas from spores on pine seedlings whereas Suillus luteus did not do so, which agreed with the results for birch.