The most assured method of producing winter wheat (Triticum aestivum L.) in the semiarid Brown soil zone of the Canadian prairies is by seeding it directly into chemical fallow. This provides adequate soil moisture at seeding, protection from low winter temperatures and results in some build-up of available soil N. However, there is usually a need for additional P. Further, to ensure that adequate levels of grain protein are attained, winter wheat may require added N fertilizer. There is also some question regarding the most effective method and depth of fertilizer placement. Two 3-yr experiments, one during a relatively dry period and the other in a moist period, were conducted on a silt loam at Swift Current, Saskatchewan, to determine grain yield and protein response of winter wheat to rates of N, P and K and to P placement. Yield response to P rates occurred when available moisture was favourable and soil-test P was low to moderate Yield response to N rates occurred in 2 of the 3 yr tested. Here too, high moisture and low to moderate levels of soil-test N resulted in a positive yield and protein response. Seed-placed P and P banded mid-row at 5- and 10-cm depth increased yields, compared with broadcast P, in one moist year but not in two dry years (exp. 1). In three moist years (exp. 2), P placement (seed-placed and mid-row-banded at 2.5- and 5-cm depths) had no effect on yield. Rate of K (0 vs. 100 kg K2O ha−1) did not influence grain yield or protein in three moist years because soil-test K was high. In three moist years (exp. 2), grain protein concentration was positively influenced by N rate, but neither P rate nor method of placement affected protein. We recommend that producers who choose to grow winter wheat on chemical fallow in the Brown soil zone apply N and P fertilizer based on a soil test. Further, they should either seed-place the P or band it at a shallow depth (about 5 cm). Key words: Grain yields, protein, fertilizer response, potassium