Abstract A common protocol employed for the preparation of conidia employs flooding a fungal colony grown on semi-solid media under optimum conditions with an aqueous solution. In contrast, conidia produced in a natural environment are usually not hydrated when disseminated in air and can be produced under water stress. In order to simulate the latter conditions, cultures were grown at different water activities and conidia were dry-harvested on the lid by turning the dishes upside-down then gently tapping the bottom of the box. This study aimed at assessing the effect of the preparation of fungal conidia on their inactivation by ethanol vapours. Firstly ethanol vapours (either 0.30 or 0.45 kPa) were applied to conidia obtained from the standardised protocol and to dry-harvested conidia for some species of Penicillium. While all dry-harvested conidia remained viable after 24 h of treatment, about 1.0, 3.5 and 2.5 log 10 reductions were observed for hydrated conidia of Penicillium chrysogenum, Penicillium digitatum and Penicillium italicum respectively. Secondly ethanol vapours (0.67 kPa) were applied to dry-harvested conidia obtained from cultures grown at 0.99 a w and at reduced water activities. For all species, the susceptibility to ethanol vapours of conidia obtained at 0.99 a w was significantly greater than that of conidia obtained at reduced water activities. Conidia produced in a natural environment under non-optimal conditions would be much more resistant to ethanol vapours than those produced in the laboratory. This phenomenon may be due to a reduced intracellular water activity of dry-harvested conidia.