Abstract A major objective of tree improvement programs is to identify genotypes that will perform well in operational deployments. Relatively little is known, however, about how the competitive environment affects performance in different types of deployments. We tested whether the genetic composition and density of deployments affect the performance of full-sib families of Douglas-fir (Pseudotsuga meniziesii), and whether traits related to crown morphology could help to explain differences in family performance under competition. Seedlings from eight families were planted in pure-family and three mixed-family composition treatments at high, medium, and low densities (11,954, 2988, and 747 trees·ha−1, respectively). Height (HT), diameter at breast height (DBH), and volume·ha−1 (VOLHA) were measured at ages 8 and 15 years. Significant differences were found among composition treatments in all traits other than VOLHA8 and significant interactions between composition and density treatments were found for all traits at age 15 years. Family ranks for DBH15 and VOLHA15 in pure-family treatments changed considerably among densities, but ranks were more stable for HT15. The performance of two mixed-family treatments differed significantly from the average performances of the same families in pure-family treatments for several traits. Differences in DBH15 among families in high-density, pure-family treatments could be explained in part by differences in crown morphology, with better performance among families with relatively narrow crowns, stout branches, and high leaf area relative to branch length. Our results suggest that the competitive environment has a considerable effect on family performance, and that incorporating crown morphology traits into selection criteria in tree improvement programs may lead to greater productivity of Douglas-fir.