Knowledge of the quality of fruits and the genetic variability among genotypes of acerola is important to optimize the use of genetic resources in improvement programs. The objective of the present study was to assess physico-chemical characteristics to determine genetic divergence among 24 genotypes of acerola. Sixty ripe fruits were collected per plant in a commercial orchard in the municipality of Jequitibá-MG. They were assessed in two production seasons with regard to their physical, chemical, and physico-chemical characteristics. Multivariate analysis of the genetic divergence among the genotypes was performed using principal component analysis, Tocher’s clustering method, and the nearest-neighbor hierarchical method, using the standardized mean Euclidean distance as the measure of dissimilarity. Genotypes 13, 20, and 23 exhibited means that were higher than the overall mean values for the main characteristics (fruit mass, ratio, and vitamin C content). Genotypes 3, 6, 8, 21, and 24 stood out with regard to fruit mass and ratio, i.e., they were larger and sweeter. Genotype 16 stood out with regard to fruit mass and vitamin C content. Results obtained using the methods were comparable, and the methods were effective in discriminating between the genotypes. This indicated variability, with potential for use in improvement programs. Based on the mean values for the chemical and physical characteristics of fruits and considering the genetic divergence among the genotypes, it was possible to suggested nine crosses to generate a future improved population, namely between genotype 11 and genotypes 3, 6, 8, 13, 16, 20, 21, 23, and 24.Knowledge of the quality of fruits and the genetic variability among genotypes of acerola is important to optimize the use of genetic resources in improvement programs. The objective of the present study was to assess physico-chemical characteristics to determine genetic divergence among 24 genotypes of acerola. Sixty ripe fruits were collected per plant in a commercial orchard in the municipality of Jequitibá-MG. They were assessed in two production seasons with regard to their physical, chemical, and physico-chemical characteristics. Multivariate analysis of the genetic divergence among the genotypes was performed using principal component analysis, Tocher’s clustering method, and the nearest-neighbor hierarchical method, using the standardized mean Euclidean distance as the measure of dissimilarity. Genotypes 13, 20, and 23 exhibited means that were higher than the overall mean values for the main characteristics (fruit mass, ratio, and vitamin C content). Genotypes 3, 6, 8, 21, and 24 stood out with regard to fruit mass and ratio, i.e., they were larger and sweeter. Genotype 16 stood out with regard to fruit mass and vitamin C content. Results obtained using the methods were comparable, and the methods were effective in discriminating between the genotypes. This indicated variability, with potential for use in improvement programs. Based on the mean values for the chemical and physical characteristics of fruits and considering the genetic divergence among the genotypes, it was possible to suggested nine crosses to generate a future improved population, namely between genotype 11 and genotypes 3, 6, 8, 13, 16, 20, 21, 23, and 24.