The compositional nutrient diagnosis (CND) method considers the multiple relationships among nutrients and has been proposed to evaluate the nutritional status of plants in place of the univariate and bivariate methods. As it is mathematically based and considers the interactions among all nutrients at the same time, it avoids the errors and trends observed in the calculations of other methods estimating nutritional status, enabling a greater relationship with productivity. The objective of this study was to obtain the CND norms for high-yielding populations of potato crops. For this, 587 samples were used from 21 experimental areas in the state of São Paulo, Brazil to correlate the leaf nutrient contents and the yields of potato crops. Crops with yields higher than 48,993.24 kg ha−1 were considered to have high yields, and the Mahalanobis distance separated the balanced samples from the nutritionally unbalanced ones. Thus, the CND-ilr method generated the norms and classified the 587 samples as nutritionally balanced with a high yield (5% of the total), nutritionally unbalanced with a low yield (92%), nutritionally unbalanced with a high yield (0.3%), or nutritionally balanced with a low yield (2.7%), with accuracy, sensitivity, specificity, NPV, and PPV scores of 96.9, 97.1, 93.6, 64.4, and 99.6%, respectively.
