SummaryThe contributions by scion and by rootstock to a number of aspects of the growth and form of four-year-old deblossomed compound trees have illustrated that major contrasts may occur between the influence of a given clone when present as a scion to that when present as a rootstock. Nevertheless, contributions by scion and by rootstock, assessed upon all combinations as scion and as rootstock of all clones in the experiment, are largely additive, mean squares for scion/rootstock interaction being often statistically insignificant or, relatively, quite unimportant. Such conclusions arise, with the clones examined, in the cases of apple, quince and cherry.