Smoke and smoke-like analogues play a key role in the release of deep seed dormancy for a wide variety of geosporous (and three bradysporous) Australian plant species. Heat and ash were thought to be the primary cues for release of dormancy and subsequent germination. It is now known that smoke, applied as an aerosol to unburnt bushland soil containing seeds of geosporous species, or to seeds buried in soil in germination trays, is a key agent in the promotion of germination. In natural bushland sites, application of smoke increased germination 42- (Banksia woodland) to 48-fold (jarrah [Eucalyptus marginata]-tuart woodland) greater than the control. In 37 species intensively examined in the Banksia woodland, only three species (a weedy annual, a legume and Hibbertia amplexicaulis) showed a more promotive effect of the heat/ash of the wildfire than smoke applied to adjacent unburnt vegetation. The action of smoke in releasing dormancy is complex and varies between species -- in some cases involving permeation of lipoidal deposits in the sub-testa to denaturing of inhibitor compounds in the seed endosperm and/or embryo. The active principle in smoke is water soluble, acts at extremely low concentrations (nanograms) and is deposited in the soil surface following fire. Rainfall solubilizes the germination-active chemical species, transmitting the chemicals to the soil seed bank in the first 25-50 mm of precipitation. Smoke is therefore a reliable and precise measure of the action of fire in the landscape for germination of geosporous species. Detailed studies indicate that smoke provides a valuable tool for auditing of the soil seed bank of many important understorey species, particularly fire sensitive (seeder) species. In addition, smoke can act as a surrogate for fire for understanding the dynamics of season of fire on recruitment and survival of species. Importantly, germinant number, diversity and survival have now been shown to be intimately linked to the season of smoke (fire) application. This chapter explores commonly held principles regarding the role of fire in ecosystem processes as they relate to germination, growth and flowering phases in some plant groups from south-west Western Australia. Fire has often been interpreted in terms of the perceived adaptations by plants to fire. Here we attempt to show the role of fire as an agent in ecosystem function for which there is still much to be learnt -- prosecuting the need for further research on the role of fire in native ecosystem function and how biological complexity (the combined interaction of function and diversity) is maintained in the absence of fire. © 2003 Buckhuys Publishers, Leiden, The Netherlands.