Genetic improvement of Douglas-fir in New Zealand was initiated in 1955 with large provenance trials established in the late 1950’s. These trials illustrated that material from the coast of Oregon and California grew faster than other provenances tested. Further collections were made to evaluate provenance and familylevel performance from these two areas, and in 1996 additional trials were established at four low-altitude sites across New Zealand. Genotype×environment (GxE) interaction among these sites was found to be important for diameter at breast height (DBH), less important for stem straightness and malformation and not important for outerwood acoustic velocity (a surrogate for wood stiffness). Heritabilities were low to moderate for all growth traits, and very low for malformation. Heritability for needle cast due to Swiss needle cast, measured as needle retention on the one site where infection was relatively high, was moderate at 0.37, and was likely a major factor creating GxE interactions for growth among sites. The heritability of wood acoustic velocity was moderate to high at individual sites (0.26-0.74) and across sites (0.49). Individual- trait selection revealed the potential for good genetic gains to be made when selecting the top 20 families for diameter growth (an average of 10.7%), straightness (an average of 11.5%) and acoustic velocity (an average of 7.0%). When we examined predicted genetic gains while selecting for needle retention and/or DBH, we found that selecting for needle cast at the affected site did not compromise DBH gains at that site. Selecting for genotypes with low needle cast at the affected site did, however, reduce gains for DBH estimated across all sites. In order to maximise gains across the current Douglas-fir growing estate, a division of growing sites between those known and predicted to be affected by needle cast and those not affected would seem appropriate. This is particularly relevant given recent climate modelling work suggests that Swiss needle cast will become more important in the South Island, and even more destructive in the North Island of New Zealand. We suggest addressing differences in site through the development of separate deployment populations.