Six hundred 7-year-old Eucalyptusnitens (Dean & Maid.) trees from 50 open-pollinated families were measured for wood density and Pilodyn penetration across two contrasting sites in eastern Victoria, Australia. Eight Pilodyn observations, two from each of four aspects, were made at a height of 1.3 m. Density was measured on whole disks cut from 1.3 m. Heritability of Pilodyn penetration and disk density at 1.3 m were 0.60 and 0.73, respectively. Phenotypic and genetic correlations between Pilodyn penetration and density at 1.3 m were −0.59 and −0.92, respectively. The high repeatability of Pilodyn penetration (0.90) suggests that only two observations per tree would be required for indirect selection of density. Direct index selection for density gave an expected 13% gain (assuming a selection intensity of 1%), compared with a 11% gain by using indirect Pilodyn selection, a selection efficiency of 84%. However, Pilodyn sampling is faster, cheaper, and not destructive, thus resulting in overall higher expected gains for selection of trees or culling of seedling seed orchards in comparison with the more destructive direct assessment of density.