This study aimed to evaluate the efficiency of the parameters fresh leaf mass and stomatal density in the estimation of ploidy and pre-selection of putative banana autotetraploids. Young plants of the diploid cultivar Ouro (AA), previously subjected to in vitro polyploidization with the antimitotic amiprofos-methyl - APM (0, 10, 20, 30, 40, and 60 µM L-1) and colchicine (2.5 mM L-1) were evaluated for survival, height, number of leaves, pseudostem diameter, leaf disc fresh mass, and stomatal density. Ploidy was determined by flow cytometry in a random sample of 200 plants to find the relationship between the genomic content and the analyzed variables. Spearman’s correlation analysis showed a strong correlation (0.84**) between leaf disc fresh mass and tetraploid plants, as well as an absence of significant correlation between stomatal density and ploidy levels. Thus, the pre-selection of plants was carried out using a reference value of leaf disc fresh mass and resulted in 688 plants pre-selected as putative tetraploids. Among them, 318 were confirmed as tetraploids, 291 as diploids, and 79 as mixoploids after flow cytometry analysis. The pre-selection efficiency was 46.2% of tetraploid plants from the total pre-selected. The results indicate that the use of leaf disc fresh mass is a simple, practical, and promising method to estimate ploidy when a high number of plants is obtained in in vitro banana polyploidization studies, reducing the total number of plants to have confirmed ploidy.