Pinus contorta Dougl. (Lodgepole pine) and Picea sitchensis (Bong.) Carr. (Sitka spruce) account for c. 50 and 20%, respectively, of the forest trees currently being planted by the Forestry Commission in Britain. Both have extensive natural ranges in western North America, and provenances (seed populations) collected in different geographic zones differ in morphology and phenologyPinus contorta (Critchfield 1957; Jeffers & Black 1963; Roche 1966; Hagner 1970); Picea sitchensis (Burley 1965; Daubenmire 1968; Roche 1969). The variation in yield and quality on sites in Britain is currently being assessed and exploited by Forestry Commission staff (Lines 1965, 1966, 1968; Faulkner, Herbert & Fletcher 1969) who made provenance trials available for this study. Rapid leader growth and stem volume increase is required, but with few short branches in a narrow crown to minimize the number of wood knots and risk of wind damage (e.g. causing 'basal bowing' or 'butt sweep' in Pinus contorta, Edwards, Atterson & Howell 1963; Lines & Booth 1972). However, stem volume is proportional to crown volume in young trees (Assman 1970), and therefore a compromise in branch development may be sought to ensure a rapid early build-up in photosynthetic tissue, light interception and suppression of ground vegetation, which seem to confer lasting yield advantages (Lines 1971; Lines & Booth 1972). The size and form of the individual tree crowns depend on (a) the amount of leader extension growth, and (b) the numbers, sizes and disposition of the branches. Seedling studies have usually shown that leader growth of Picea sitchensis, and in part, of Pinus contorta, tends to increase with decrease in latitude or elevation of seed origin, reflecting ecotypic or clinal differences in photoperiodic and temperature response-P. contorta (Critchfield 1957; Sweet & Wareing 1968; Hagner 1970); Picea sitchensis (Aldhous 1962; Burley 1966a, b; Lines & Mitchell 1966). Provenance differences in branch development, however, have not been described in sufficient detail to identify the morphological characters responsible for well-known differences in crown form. This study first tested the assumption that the length of needle-bearing shoot determined the amount of needle tissue produced, and that this was correlated with stem volume production. This was necessary because, in both Pinus contorta and Picea sitchensis, provenance differences are known in needle size, closeness, display (Critchfield 1957; Jeffers & Black 1963; Lamb 1970; Burley 1965) and photosynthetic rate (Sweet & Wareing 1968; Ludlow & Jarvis 1971). Following this, the lengths of all the shoots on young trees were measured to trace how some provenances had built-up much more shoot than others. A simple model was used to establish how much these differences stemmed from differences in (a) leader growth, (b) the length of each branch relative to its neighbours and the leader, and (c) lateral bud production.