Studies were made in autumn 1985 of the forking defects of 1983 and 1984 shoots of trees in a 9-yr-old progeny test. Lammas growth appeared to be the main factor explaining forking. A score for branching angle was useful in evaluating causes of forking other than Lammas growth. These 2 characters were good predictors of forking, with high variability and heritability. For genetic improvement of the species, selection should be for total ht. and small branching angle and against Lammas growth.