The performance of hybrid larch (Larix × eurolepis) stands may be altered by the fact that the seed lots produced in hybridization seed orchards always contain a proportion of non-hybrid seeds. We studied the effect of such hybrid/non-hybrid mixtures on the productivity and quality of young plantations, through a three-step process: determination of tree species identity with cytoplasmic DNA markers, comparison of hybrid and non-hybrid tree performance, and finally, assessment of thinning impact on hybrid percentage. Overall, we analyzed progenies from eight commercial seed orchards at three sites. Huge variations in hybrid percentage were found among orchard progenies. Globally, the results concerning survival, susceptibility to drought and to Meria laricis, growth, stem form and wood stiffness were clearly in favor of hybrids. In particular, they grew faster than the non-hybrid trees and exhibited a more homogeneous growth than the mixed progenies. Hybrid superiority was highest for orchards composed of a single maternal clone of a species which was poorly adapted to the planting site. Hybrid percentage was enhanced by successive thinnings. However, a serious loss of income is expected when planting progenies characterized by low initial hybrid percentage and high hybrid superiority. Moreover, variations in hybrid percentage proved to be problematic in seed orchard testing as seed orchard ranking may change depending on whether it is based on the performance of all trees or hybrids only. These variations reduce the reliability of the recommendations typically made to forest owners. For commercial and silvicultural reasons, we suggest setting a minimal threshold level of 60–70 % for hybrid percentage in marketed forest reproductive materials.