AbstractKnowledge about the diversity in the morphological characteristics of robusta coffee (Coffea canephora Pierre ex Froehner) flowers is fundamental to understanding reproduction processes, interpret pollination‐related mechanisms, and contribute to breeding studies. The objective of this study was to analyze and characterize morphological variability of flower traits among robusta coffee cultivar Conilon genotypes of high agronomic performance. The experiment was arranged in a randomized block, split‐plot design. The main plot corresponded to six Conilon coffee genotypes and the subplots to 2 yr of evaluation (2018 and 2019). Eight floral traits were measured and analyzed through analysis of variance by the F test (<.01), Tocher's optimization method and hierarchical Unweighted Pair Group Method using Arithmetic Averages (UPGMA) clustering. The relative importance of each character was also estimated to predict genetic diversity, and the degree of trait interrelation was assessed by Spearman's correlation. Genotype Pirata had outstanding floral morphological structure and was grouped separately by Tocher and UPGMA clustering. Style length was the major contributor to genetic diversity, followed by the lengths of the corolla tube and stigmatic lobes. The flower trait characteristics were positively correlated, and corolla tube length and style length were most strongly correlated. These results contribute to other studies related to the interpretation of the mechanisms related to pollination and reproduction of this species, and also outline diversity of genetic resources of interest to plant‐breeding programs.