Predicting yield at an early growth stage allows one to optimize the amount and timing of nitrogen fertilization to maximize crop yield. The aim of the study was to determine how different nitrogen doses would affect the nutrition of two maize varieties (Zea mays L.) with this element and their yields, and to demonstrate the potential of using the SPAD index as an early-stage predictor of dry matter yield (DMY) of maize biomass. The field experiment, with different pre- and post-sown N fertilizer doses, was conducted for two years in Poland. The two-factor experiment was set up in a randomized block design with three field replicates. There were four mineral fertilization variants: (I) 120 kg N·ha−1 without P and K fertilization, (II) 120 kg N·ha−1 + PK, (III) 160 kg N·ha−1 + PK, and (IV) 200 kg N·ha−1 + PK. The second-order factor was two stay-green maize varieties: Delitop (FAO 240) and Magitop (FAO 270). The mean values of the SPAD index of maize depended on the nitrogen dose. When the dose was increased from 120 kg·ha−1 to 200 kg·ha−1 with constant P and K fertilization, the value of the leaf greenness index increased significantly. The plants of both maize varieties fertilized with nitrogen at a dose of 120 kg N·ha−1 without P and K were characterized by the lowest values of the SPAD index, regardless of the year of the experiment. When the nitrogen dose was supplemented with P and K, the plants’ nutrition with nitrogen improved significantly. The values of the SPAD index in the maize plants of the variety Magitop were significantly greater than in the plants of the variety Delitop. The coefficients of correlation calculated for both maize varieties were highly significant at all measurement dates. Therefore, it is possible to conclude that the SPAD leaf greenness index can be used as an important, reliable, and non-invasive indicator for maize yield prediction. Nutrition of maize plants with nitrogen proved to be the main determinant of variability of the biomass yield. For this reason, it was the basis for the verification of individual mineral fertilization variants. The lack of phosphorus and potassium in the dose of mineral fertilizer limited the yield regardless of the variety, as compared to the variant that was optimally balanced with the nitrogen dose.