Understanding the nutrient dynamics of pastures is essential to their profitable and sustainable management.  Tall fescue [Schendonorus phoenix (Scop.) Holub.] is the predominant forage species in Virginia pasturelands.  Although tall fescue pasture is common, little research has attempted to document how soil and herbage nutrient concentrations change through time.  This thesis summarizes two studies conducted within the context of a larger grazing systems project near Steele's Tavern, VA.  The objectives were to: (1) examine temporal changes in plant available soil nutrient concentrations in four grazing systems, (2) determine how hay feeding and use of improved forages affected soil and herbage nutrient concentrations (3) examine the relationship between and variability within soil and herbage nutrient concentrations, (4) analyze the seasonal variation in herbage mineral concentration with regard to beef cattle requirements, and (5) create a statistical model to predict variation in herbage mineral concentration across the growing season.  Analysis of plant and soil nutrients through 5 years of grazing produced several important findings.  Soil pH, P, and Ca, Mg, and B declined through five years of grazing.  Higher concentrations of herbage N and K and soil P, K, Fe, Zn, and Cu were measured in hay feeding paddocks.  Herbage nutrient concentrations showed less variability in P and K than did soil test results.  Fertility testing in pastures is important to monitor changing nutrient concentrations, and this study showed that herbage analysis may provide a more stable and accurate assessment of pasture fertility than soil testing.  Pasture herbage, grown without fertilization, contained sufficient concentrations of macronutrients to meet the requirements of dry beef cows through the growing season and to meet the requirements of lactating beef cows in April.  A model was developed using soil moisture and relative humidity that predicted (R2 = 0.75) variation in herbage mineral concentration throughout the growing season.  As described in this thesis, use of modeling to predict nutrient dynamics in pasture could allow for more efficient mineral supplementation strategies that lead to improved profitability, nutrient retention, and livestock health. / Master of Science