Within the framework of the eucalyptus breeding programme in the Congo, two retrospective tests were conducted using mature clones in the field and young cuttings under nursery conditions with two hybrids: 13 clones of Eucalyptus tereticornis* Eucalyptus grandis for the test TC 82-1B and 17 clones of Eucalyptus urophylla* Eucalyptus grandis for the test TC 86-10. Growth, morphological and biomass traits were measured during the nursery phase and growth traits were measured in the field trials. Analysis of variance showed that the clones were significantly different for most of the traits at the two stages. However, for the clonal test TC 82-1B, all the coefficients of correlation between juvenile and mature traits were very low (around 0,3) and not significantly different from zero. For the clonal test TC86-10, several coefficients were significantly different from zero but their value was small, around 0,5. This weak correlation was explained by the weak control of growth by genes at the young cutting stage, the morphology of cuttings, the non-optimal testing conditions, the high clone by environment interaction and the reduced genetic base of the clonal population. For clonal test TC86-10, multitrait selection, using the association of total growth and growth increment or biomass traits increases the coefficient of determination. These initial results show that early selection of Eucalyptus clones could be used to reject of around 30% of the poorer growing clones at the mature stage.