The symbiotic relationship between Rhizobium bacteria and legume host plants has long been used by humans to produce forage legumes on N-limited soils. However, increased producer acceptance and widespread use depend on several factors, including establishment success of annual and perennial species. Crimson clover (Trifolium incarnatum L.) and white clover (Trifolium repens L.) were chosen as model crops and either no-till drilled or broadcast seeded at two rates into a bermdudagrass (Cynodon dactylon L.) sward and grazed by cattle for 5 d either before or after establishment in a randomized complete block design with three replications. Maximum counts observed in 2009 were 215 seedlings/m² and similar for crimson and white clover. Seedling numbers in broadcasted white clover plots that were ‘grazed after'at a high seeding rate were elevated (P<0.05) compared with ‘grazed before' plots in 2009, but this effect was not observed elsewhere and did not reappear during 2010, making a consistent cattle hoof-action effect for providing increased soil-seed contact unlikely. Yields of above-ground biomass collected in spring (crimson and white clover) and autumn (white clover only) generally followed seedling count patterns. Dry matter yields of crimson clover in early spring of 2009 ranged between 1.2 Mg/ha (no-till drill high seeding rate) and 200 kg/ha (broadcasting low seeding rate); those of white clover ranged between 210 and 6 kg/ha. In autumn of 2009, white clover DM yields ranged between 180 and 27 kg/ha. An economic analysis is currently underway to determine which establishment method is the most cost-effective in terms of spring seedling counts and biomass.