ABSTRACT This study aimed to select regression equations based on the correlation between chlorophyll index and leaf nutrient contents, for real-time prediction of the nutritional status of 'Prata' banana. Six cultivars of 'Prata' banana were used as treatments, with five replicates and four plants per plot, arranged in a completely randomized design. Nutrient levels were evaluated based on laboratorial analysis and chlorophyll indices using a portable chlorophyll meter, in the third leaf from the apex to the base. Data were subjected to analysis of variance, calculating the correlations between leaf nutrient contents and chlorophyll indices, and the regression equations were adjusted to the associations that were significant and with greater magnitude. The selected models estimate leaf nutrient contents and allow a real-time, low-cost reliable prediction of the nutritional status of 'Prata' banana. It is an auxiliary technique that, combined with leaf and soil analysis, contributes to a more precise and secure nutritional evaluation.