The effects of test environments (dry versus wet) on estimation of genetic parameters in seedling traits were studied in 160 open-pollinated families of Douglas-fir (Pseudotsuga menziesii (Mirb.) Franco) from southwestern Oregon. Seedlings from four populations were grown in two test environments for two growing seasons. Between test environments, -9 bars of water potential difference were created in both growing seasons. Estimated genetic variances in most growth and phenology traits were considerably higher for seedlings grown in the wet environment than for those in the dry. Estimated genetic correlations between the same traits measured in different test environments indicated that most seedling traits studied for two growing seasons are genetically stable in both environments-i.e., suggesting that genotype by environment interaction in these traits are weak. However, the effect of test environment on estimation of genetic parameters in seedling traits, especially in adaptive seedling traits, should be evaluated very carefully when early evaluation of genetic entries is practiced in Douglas-fir since these traits (budburst timing, lammas growth and free growth) appear to be plastic traits.