ABSTRACT Peppers belong to the Solanaceae family and present a wide genetic variability that may be accessed through phenotypic and genotypic traits. This work aimed to study the genetic divergence in an ornamental pepper F3 population by multivariate methods based on the individual and simultaneous analyses of qualitative, quantitative, and molecular data. The work was developed in the Center of Agrarian Sciences (CCA) of the Universidade Federal da Paraíba (UFPB), Paraíba state, Brazil. 44 progenies from an F3 generation were used and the characterization of 30 qualitative traits, 16 quantitative traits, and 18 pairs of microsatellite primers was performed. Individual and simultaneous analyses of the variables were performed by using the clustering method of Ward’s algorithm. Tocher’s method was used based on distance matrixes. A non-metric multidimensional scaling was applied (nMDS) for the graphic representation of the distance matrixes in the bidimensional space. Data analysis was efficient to separate the genotypes into distinct groups. There is genetic variability among the individuals of the C. annuum F3 population, verified by individual and joint trait measurements. The three clustering methods were efficient to represent the genetic distance between individuals of ornamental pepper plants. Individuals 1, 2, 7, 8, 10, 11, 21, 27, 29, 35, and 38 are indicated to advance generation and continue the breeding program.