The effects of temperature and water availability on growth and interactions between fumonisin-producing isolates of Fusarium moniliforme and F. proliferatum and seven other fungi from maize grain were determined in vitro. The type of interaction and index of dominance (ID) between species were markedly influenced by temperature and aw. Generally, F. moniliforme and F. proliferatum were very competitive and dominant against the Penicillium spp. and A. flavus. They were in turn dominated by A. niger, but mutually antagonistic when paired with F. graminearum and A. ochraceus. Under slightly drier conditions (<0·98 aw) A. ochraceus became more competitive and dominant over the fumonisin-producing species. A. flavus was dominant only at 30°C and <0·96 aw. F. moniliforme and F. proliferatum demonstrated dominance against all species over a range of temperatures and 0·994 to 0·96 aw. At lower aw levels they were less competitive. The growth rate of the two fumonisin-producing species was significantly reduced by F. graminearum, regardless of aw. F. moniliforme and F. proliferatum reduced growth of Penicillium and Aspergillus spp., especially at >0·96 aw. At <0·96 aw, growth of these species was unaffected. Using Biolog plates the effect of aw and temperature on utilization patterns of carbon sources in maize were evaluated for the first time. The niche overlap indices relative to F. moniliforme and F. proliferatum were determined and compared with that of each interacting species. NOIs for F. moniliforme and F. proliferatum were >0·90 at >0·96 aw and 25 and 30°, indicative of co-existence with other species. Most of species had NOIs >0·90, except in some cases when paired with F. moniliforme, where NOIs <0·80 suggested the occupation of different niches. Although there was no significant correlation between the ID and NOI methods both suggested that the niche overlap between species was in a state of flux and significantly influenced by both temperature and water availability. This suggests that interpretation of ID, or NOIs carried out under one set of environmental conditions may be misleading when considering interactions between species and also where screening for biocontrol potential is being considered.