Increasing atmospheric carbon dioxide concentrations present a novel resource condition for plant communities. In order to understand and predict how plant community structure and function may be altered in a high CO2world, we need to understand how interactions among neighbouring plants within a community will alter the growth and reproduction of component species. Because CO2 is readily diffusible, plants have little influence on the CO2 acquisition of their neighbours, except within particularly dense canopies. Thus, plants seldom compete directly for CO2. Rather, CO2 availability is likely to alter plant-plant interactions indirectly through its effects on plant growth and competition for other resources. As a consequence, competitive outcome under elevated CO2 atmospheres within even simple systems is not easy to predict. For example, under some conditions, C4 species in competitive assemblages have improved competitive ability relative to C3 competitors as a result of CO2 enrichment, contrary to expectations based on their photosynthetic pathways. It is now clear that individually grown plants can differ substantially from those within mono- or multispecific stands in response to CO2 enrichment. At present, our understanding of how stands of interacting plants modify the availability of CO2 and other resources is incomplete. We urgently need information about how elevated CO2 atmospheres influence stand formation and population dynamics, specifically with regard to the identities, numbers, sizes and reproductive fitnesses of individuals within single and multiple species stands, if we are to make multi-generational predictions concerning the fate of populations and communities in an elevated CO2 world.