I propose to discuss some of the roles that fungi may play in terrestrial ecosystems because it is clear from recent work that the magnitude of their intervention in nutrient and energy cycling may be very great. Although fungi are highly diverse and intervene in many different ways, many of the methods at present in use to describe their activities, even in general terms, are at best very arduous and at worst extremely liable to error. I shall therefore first consider a few important facets of the biology of fungi which are necessary both to an appreciation of their ecological roles and to the elaboration of satisfactory methods for their detailed investigation. The peculiar difficulty of the study of fungi in natural habitats originates from the microscopic dimensions of their vegetative bodies. Their hyphae are a few microns in diameter and penetrate into and through their substrates. Except in the broadest sense, they cannot be identified by vegetative characters. Field identification is, therefore, possible only of those which produce macroscopic fruiting bodies and, even with these, there is a limit to the conclusions which can be reached about their ecological activities by observing the distribution and frequency of their reproductive stages. The methods of studying fungi in ecological situations must, therefore, combine procedures of direct observation with techniques of isolation, and every effort made to distinguish those which are present as active vegetative mycelia from those present as dormant spores. The difficulties are great and it has not yet been possible to decide, with certainty, for any complex ecological site such as a soil horizon, what kinds of fungi are present in an active state let alone in what proportion the different species occur. Much of the information till recently was acquired in particular as a by-product of plant pathology. One of the important reasons for this was that the problems attacked had each the limited objective of explaining the relationship of a particular disease organism to its host substrate and to other organisms occurring around or upon that substrate. Indeed, it was in the investigation of plant pathogenic fungi that biological antagonism as distinct from competition was first discovered. In consequence of this kind of by-product, the study of the ecology of pathogens, especially in the soil, has excited interest in wider aspects of fungal ecology (see Baker & Snyder 1965). Detailed investigations of rhizosphere organisms and of populations of different species living together on root surfaces arose from this source and so led to ecological classifications of fungi on such bases as ability to compete as saprophytes or their restriction to particular sites or substrates in the soil, The pathogenic fungi whether unspecialized necrotrophs or obligate biotrophs were seen each to have its own ecological niche and to be subject to interaction and competition with other organisms and to be controlled in development by the condition of the hosts. Host and pathogen are indeed subject to joint and separate