The mass of nutrients accumulated in the vegetative parts of winter wheat (WW) in the period from the beginning of booting to the full flowering stage (Critical Cereal Window, CCW) allows for the reliable prediction of the grain yield (GY) and its components, and the grain protein content (GPC) and its yield. This hypothesis was verified in a one-factor field experiment carried out in the 2013/2014, 2014/2015, and 2015/2016 growing seasons. The field experiment included seven nitrogen-fertilized variants: 0, 40, 80, 120, 160, 200, and 240 kg N ha−1. The N, P, K, Ca, Mg, Fe, Mn, Zn, and Cu content in wheat vegetative parts (leaves, stems) was determined in two growth stages: (i) beginning of booting (BBCH 40) and (ii) full flowering (BBCH 65). We examined the response of eight WW traits (ear biomass at BBCH 65, EAB; grain yield, GY; grain protein content, GPC; grain protein yield, GPY; canopy ear density, CED; number of grains per ear, GE; number of grains per m−2—canopy grain density, CGD; and thousand grain weight, TGW) to the amount of a given nutrient accumulated in the given vegetative part of WW before flowering. The average GY was very high and ranged from 7.2 t ha−1 in 2016 to 11.3 t ha−1 in 2015. The mass of ears in the full flowering stage was highest in 2016, a year with the lowest GY. The highest N mass in leaves was also recorded in 2016. Only the biomass of the stems at the BBCH 65 stage was the highest in 2015, the year with the highest yield. Despite this variability, 99% of GY variability was explained by the interaction of CGD and TGW. Based on the analyses performed, it can be concluded that in the case of large yields of winter wheat, GE is a critical yield component that determines the CGD, and in consequence the GY. The leaf nutrient mass at the BBCH 40 stage was a reliable predictor of the GPC (R2 = 0.93), GPY (0.92), GE (0.84), and CED (0.76). The prediction of the GY (0.89), CGD (0.90), and TGW (0.89) was most reliable based on the leaf nutrient mass at the BBCH 65 stage. The best EAB prediction was obtained based on the mass of nutrients in WW stems at the BBCH 65 stage. The magnesium accumulated in WW parts turned out to be, with the exception of TGW, a key predictor of the examined traits. In the case of the TGW, the main predictor was Ca. The effect of Mg on the tested WW traits most often occurred in cooperation with other nutrients. Its presence in the developed stepwise regression models varied depending on the plant part and the WW trait. The most common nutrients accompanying Mg were micronutrients, while Zn, Fe, Mn, and Ca were the most common macronutrients accompanying Mg. Despite the apparently small impact of N, its yield-forming role was indirect. Excessive N accumulation in leaves in relation to its mass in stems, which appeared in the full flowering phase, positively impacted the EAB and GPC, but negatively affected the GE. Increasing the LE/ST ratio for both Mg and Ca resulted in a better formation of the yield components, which, consequently, led to a higher yield. This study clearly showed that nutritional control of WW during the CCW should focus on nutrients controlling N action.