The results of earlier axenic and dual cultures of Dactylorhiza purpurella and its symbiont indicated that the fungus may provide the orchid with some, if not all, of the vitamins in yeast extract. Here, the fungus was studied for such production in axenic liquid still cultures on mineral–sugar media, using bioassays and chemical methods. The latter proved too insensitive. Bioassays were considerably more sensitive and, after various modifications of the techniques, gave consistent results. They showed that about 33 parts per billion (ppb) nicotinic acid and (or) nicotinamide ('niacin'), 1.5 ppb thiamine, but no pantothenic acid, were readily secreted in the culture medium by the fungus. These were, respectively, at about 10% and 50% of their concentrations in the medium used for axenic orchid cultures. 'Niacin' production was studied further. There was no real correlation with dry weights. Production was greatest with 1% dextrose (vs. 2%), during the active growth phase, and on a medium containing ammonium and nitrate vs. nitrate alone as nitrogen source. On that medium, greatest yield was under restricted aeration, and conversely for the nitrate medium. Ten micrograms 'niacin' per gram dry mycelium was released into the filtrate, and 14 μg, i.e. 40% more, was retained in the mycelium. Ammonium seems more important in the synthesis of 'niacin' than of amino acids. These findings may be of great significance in relation to symbiotic associations in dual cultures and in nature. They indicate that orchids would derive more benefit from active endophytic fungi than upon their digestion.