Conidia of Blumeria graminis (DC Speer) f. sp. hordei (Marchal) were incubated for between 1min and 3h on clean (hydrophilic) and silanized (hydrophobic) glass and up to 12h on barley (Hordeum vulgare L.) leaves from which the epicuticular waxes were removed. Specimens were cryo-fixed (−190°C) and, whilst being observed on the cold stage (<−155°C) of a scanning electron microscope, conidia were displaced from their original position using a purpose built micro-manipulator to expose original sites of conidial contact on the substrata. On silanized glass, substantial pad-like deposits of conidial extracellular material (ECM) were evident within 1min of inoculation, and these increased in size over a 3h period. On clean glass, no ECM was evident after 1min, but by 90min similar but smaller pads were present. For the first time, conidial ECM was seen on cryo-fixed barley leaves. Here, it usually occurred in the form of a limited number of small, discrete deposits. These were visible within 1min of inoculation. Up to 12h, when mature appressoria had formed, these deposits usually remained discrete and their number did not increase. The size and spatial distribution of these deposits indicated that they were released from conidial surface projections that had touched the curved surfaces of leaf epidermal cells. This also appeared to be the case on planar glass surfaces, but here ECM appeared to have flowed laterally to coalesce and form the characteristic pads. The demonstration of ECM release within 1min of spore deposition on leaves supports the view that this plays a role in recognition of the precise point of contact by conidia. The geometric relationships between conidia and planar glass substrata or the curved surfaces of leaf epidermal cells can explain the different forms of ECM deposits released by conidia. On planar surfaces, more conidial surface projections touched the substratum and released liquid ECM droplets. The combined capillary force exerted by these droplets may have been sufficient to draw the spore towards the substratum causing coalescence of droplets leading to pad formation. On curved leaf cell surfaces where fewer conidial surface projections make contact, capillary force may have been insufficient with the result that droplets remained discrete.