【Objective】The aim of this study was to clarify fertilization and requirement of nitrogen (N), phosphorus (P) and potassium (K) in spring wheat production in Yellow River Irrigation Area of Ningxia (NYRIA), so as to provide a theoretical guide for rational fertilizer reduction and application. 【Method】Based on the investigation of farmers’ fertilizer application and field fertilization experiments, farmers’ yield levels and N, P, K fertilizer application of spring wheat were analyzed, and the grain yield, yield components, and N, P, K nutrient requirement were investigated. 【Result】The average spring wheat yield of farmers was (6 985±867) kg·hm -2 in NYRIA, and the percentage of high yield farmers was 82.7%. With the increase of yield, the excessive amount of N and P was decreased, and the application rates of K was insufficient. On average, the application rates of N, P and K were 294, 162 and 49 kg·hm -2, respectively, with 97.1% of farmers over applied N by 69-114 kg·hm -2, and 20.5% of farmers over applied P by 18-42 kg·hm -2, while K inputs were generally insufficient with an amount of 30-51 kg·hm -2. The N fertilization experiment showed that shoot biomass, grain yield, harvest index and grain number per ear all increased significantly at N rates of 120-240 kg·hm -2, and all reached the highest at the application of N 180 kg·hm -2. At the same time, the amount of N, P, and K absorbed by the grain also reached to the maximum value of 168.2, 23.9, and 23.2 kg·hm -2, respectively. Nitrogen application showed no significant effect on N harvest index, the average of which was 56.5%, and tended to increase P harvest index but decrease K harvest index. Nitrogen requirement reached to 45.8 kg·Mg -1 at N rate of 180 kg·hm -2, being 19.6% higher than that of no N application. P requirement decreased from 6.0 kg·Mg -1 at no N application to 5.3 kg·Mg -1 at N rate of 240 kg·hm -2, while K requirement increased from 42.6 to 49.7 kg·Mg -1. The P experiment showed that shoot biomass and grain number per ear decreased significantly with the increase of P rates, while 1 000 grain weight and harvest index increased significantly, so there was no significant difference in grain yield over P rates. Moreover, P application improved N uptake and harvest index, being 28.6% and 27.9% higher respectively than that of no P application. Also, P fertilizer could promote P to transfer to grain because that P uptake and P harvest index were increased by 15.9% and 15.2%, respectively. However, it showed no significant effect on N and P requirement, but decreased K requirement from 68.1 kg·Mg -1 at no P application to 49.7 kg·Mg -1 at P2O5 rate of 120 kg·hm -2. The K experiment showed that the application of K had no significant effects on biomass, grain yield, harvest index, ears per hectare, and the content of N, P and K in grain. While the high K fertilizer application of 75 kg·hm -2 significantly reduced the number of grains per ear, but increased 1 000 grain weight and harvest index of N, P and K. Grain number per ear decreased by 9.1%, 1 000 grain weight increased by 7.6%, and harvest indexes of N, P and K were 57.2%, 73.5% and 7.3%, respectively. When applying 60 kg·hm -2 of K2O, the demand for N, P and K reached the highest, which was 55.3, 5.5 and 57.6 kg·Mg -1, respectively, while the demand for N and K were both significantly reduced by 20.6% and 13.7% at K2O rate of 75 kg·hm -2. It was concluded that proper application of K could increase the demand of N and K, while over application of K could reduce the demand of N and K. 【Conclusion】Reducing N fertilizer, regulating P fertilizer and properly adding K fertilizer input were still the key for spring wheat fertilization in NYRIA. The N requirement of spring wheat was ranged 38.3-57.2 kg·Mg -1, and could increase by applying the moderate amount of N, P and K fertilizer. The demand for P of spring wheat was 5.1-6.0 kg·Mg -1, which tends to increase with N rates and is not influenced by P and K amount. The K requirement of spring wheat was within 42.6-68.1 kg·Mg -1, which seems to increase with N rates and decrease with P rates, and also is reduced by high amount of K application. The suitable recommended application of N fertilizer was 120-180 kg·hm -2, which was 25%-60% lower compared with N application of farmers. Application of P fertilizer at 48-96 kg·hm -2 was more conducive to stabilize the yield of spring wheat and promote the transfer of N and P to the grain, which was reduced by 40.7%-70.3% compared with the average application of farmers. When the application of K fertilizer was 0-30 kg·hm -2, it was more benefit to stabilize the yield and increase the grain quality.