Ecological risk assessment is an important step in the production and commercialisation of transgenic plants. To date, however, most risk assessment studies have been performed on crop plants, and few have considered the ecological consequences associated with genetic modification of pasture species. In this study we compared the growth, yield, population dynamics and competitive ability of transgenic Trifolium subterraneum subsp. subterraneum cv. Leura (subclover) expressing a nutritive sunflower seed albumin (ssa) gene with the equivalent non-transgenic commercial line in a glasshouse competition trial. Plants were grown in low-fertility soil typical of unimproved native southeastern Australian grasslands. We measured survivorship, seed production rate, seed germination rate, seed weight, dry weight yield and the intrinsic rate of population increase (lambda) of plants grown in mixtures and monocultures over a range of densities (250 to 2000 plants m(-2)), and also determined intragenotypic and intergenotypic competition coefficients for each line. There were no significant differences between transgenic and non-transgenic plants in any of the measured variables except survivorship; transgenic plants had a significantly lower survival rate than non-transgenic plants when grown at high densities (p<0.01). However, density-dependent effects were observed for all measured variables, and in all models plant density affected the response variables more than the presence of the transgene. Based on these results, we conclude that the ssa gene construct appears to confer no advantage to transgenic T. s. subterraneum cv. Leura growing in mixed or pure swards under the fertility and density regimes examined in the trial. Our data also suggest that transgenic subterranean clover expressing the ssa gene is unlikely to exhibit a competitive advantage over associated non-transgenic commercial cultivars when grown in dense swards in low-fertility pastures.