Legume persistence and forage and cow-calf production were evaluated on rotationally stocked pastures in either a season-long or complementary system. In 1997, six 2.02 hectare pastures were seeded with smooth bromegrass (SB), a mixture of a grazing tolerant (GT) alfalfa and SB, or a mixture of a hay-type (HT) alfalfa and SB to be used in the season-long system. Four 2.02 hectare pastures were seeded with SB on 1.21 hectare of each pasture in mixtures of either the GT or HT alfalfa cultivars and SB on .81 hectare of each pasture to be used in the complementary system. Each pasture was divided into ten paddocks and was rotationally strip-stocked at 1.98 cow-calf units/hectare for 120 and 141 days starting May 18, 1998 (year 1) and May 6, 1999 (year 2), respectively. Each year, first harvest forage was harvested as hay from the 40% of each pasture distal to the water hydrant. In years 1 and 2, the remaining 60% of each pasture was grazed for the first 44 and 54 days and 100% of each pasture was grazed on days 45 to 120 and days 55 to 141, respectively. Legume proportions in the live DM of pastures seeded with GT and HT alfalfa cultivars decreased by 70 and 55% in paddocks stocked season-long and by 60 and 42% in paddocks used for complementary stocking (alfalfa cultivar, P < .05; stocking system, P < .05) in year 1, but decreased by 72% across cultivars and stocking systems in year 2. Total (P < .08) forage masses in September of year 1 and in August of year 2 were greater in pastures in which alfalfa paddocks were stocked season-long than in those with complementary alfalfa stocking. Grazing of alfalfa in grass mixtures increased calf and total animal weight gains in comparison to grazing season-long SB, but alfalfa persistence, measured as a proportion of the live DM, was not affected by alfalfa cultivar.