A full diallel mating design (10 parents) was carried out in a Swiss stone pine panmictic population from the Carpathian Mountains. At age six, after nursery testing, the progenies were field planted to one site, using a completely randomized block design with 100 families, four replicates and a 15 tree row-plot per replication, spaced 2.5 x 2.5 m. Total and annual height of all individuals of this test was assessed at successive ages, between ages eight and 16. In addition, weight of 100 seeds and cotyledon number were considered in correlation analyses. Significant differences occurred in total height for general and specific combining ability effects. Variance components, heritabilities, genetic correlations and genetic gains on growth traits and survival at age 16 were estimated. Across the field testing period, GCA 3) variance increased with time, accounting for 65 % of plot mean phenotypic variance by age 16. In contrast, SCA variance declined as the progeny test aged accounting for only 9 % by age 16. The time trend of the four heritability estimates for total height increased with age across the testing period reaching their highest level between age 14 and 16. Genetic correlations over time for total height rose from 0.85 at age two to 0.94 at age six and then leveled off across the field test indicating that if the goal is to improve 16-year height, early selection can be considered at age six. By selecting the best 10 %, 15 % and 20 % of individuals within the best fullsib families, a genetic gain in total height of 12.4 %, 11.0 % and 9.9 % respectively, could be achieved at age 16. A higher gain can be obtained if the best general combiner parents are selected and intermated.