A study was established to evaluate the fertilizer responses of smooth bromegrass (Bromus inermis Leyss. cv. Polar) during the first 3 years of growth on a newly cleared soil (Typic Cryochrept) in interior Alaska. The fertilization rates in the N4 by P4 by K4 complete factorial experiment were N at 0, 45, 90, and 135 kg/ha, and P and K each at 0, 34, 68, and 102 kg/ha. Yield responses to N were minimal without application of the first 34 kg/ha increment of P, and a significant N by P interaction occurred each of the 3 years of the study. During establishment (1979), the first increment of N (45 kg/ha) increased growth over sixfold above the level of the nonfertilized control. Growth was improved throughout the 0 to 102 kg/ha range of P, with over an eightfold increase in total production. Potassium (0 to 102 kg/ha) also improved growth but to a lesser extent than N or P, giving just over a 150% increase in production. In 1980 and 1981, total dry matter yields from the two cuttings increased with added N from 0.26 and 0.21 Mg/ha at the 0 N rate to 2.6 and 2.0 Mg/ha with 45 kg/ha N, respectively. Comparable yields at 90 and 135 kg/ha N were 3.8 and 4.3 Mg/ha in 1980 and 3.4 and 4.7 Mg/ha in 1981. Multiple regression equations were developed to predict bromegrass yield and N uptake for the 1980 and 1981 seasons. Relatively high coefficients of determination were obtained for dry matter yield (R2 = 0.845 for 1980 and 0.857 for 1981) and for N uptake (R2 = 0.861 for 1980 and 0.872 for 1981). Linear responses to N suggest that rates above 135 kg/ha N may provide even higher yields and protein contents.