Planting of resistant clones is the main control strategy for rust caused by Austropuccinia psidii on eucalypts in Brazil. Phenotyping for resistance to rust is performed at the last stage of the breeding programme on genetic material preselected for growth and industrial traits, which therefore may include susceptible genotypes. An alternative way to increase the frequency of resistant individuals in the breeding population is to perform resistance phenotyping during the initial stages of the programme. However, the impact of this approach on other characteristics considered as priority in the selection of genetic material should be assessed. In this study, we evaluated whether early selection for resistance to rust influenced total diameter at breast height (DBH), height (TH), average annual increment (IMA) and basic wood density (BWD), which are important selection criteria in eucalypt breeding. For this, a total of 6 703 plants from 70 full-sib progenies of Eucalyptus grandis, E. urophylla and hybrids of these species were phenotyped for rust resistance before transplanting, under controlled conditions, and again at seven months after transplanting, under field conditions and natural infection. Subsequently, the breeding population was evaluated for BWD indirectly through pilodyn penetration at 24 months after planting, and for DBH, TH and MAI at 37 months. Considering the final phenotype, 25.3% individuals were classified as resistant and 74.7% were susceptible to rust. Comparative genetic analyses between resistant and susceptible individuals showed that selection for resistance, performed early, did not influence DBH, TH, MAI or BWD traits. Thus, the early selection of resistant genotypes speeds up the exclusion of susceptible genotypes, saving time and resources in the development of new commercial clones without interfering with other important traits for tree breeding.