In the Netherlands, the rose is the most important glasshouse cut flower. Approximately 50% of the glasshouse roses are combination plants, consisting of a scion cultivar and a rootstock of a different genotype. Rootstocks inducing differences in the vigour of the scion are suggested to differ in cytokinin production and export to the shoot, thereby affecting bud break and outgrowth of the scion.In the present study, the course of endogenous cytokinins in rose combination plants was determined. The high contribution of isopentenyladenine-type cytokinins in young leaves indicated that these leaves, as the roots, were capable of de novo synthesis of cytokinins. Export of cytokinins from the roots was estimated based on the assumption that the cytokinin concentration in bleeding sap is representative for the concentration in xylem sap in situ , which was experimentally verified. The concentration of zeatin riboside (ZR) in bleeding sap was shown to be correlated with bud break of axillary shoots and bottom breaks, it increased before bud break and decreased thereafter. This suggests a quantitative relationship between cytokinin export from the roots and shoot development. Growth of the scion was also influenced by environmental factors, since bud break was advanced at higher root temperatures in the range of 11 to 26°C.However, this effect could not be correlated reliably with the cytokinin export from roots to shoot. Grafting the scion Madelon on rootstocks varying in vigour revealed that the rootstock that induces earlier bud break of the scion supplies larger quantities of cytokinins to the shoot. As a consequence, the concentrations of cytokinins in bleeding sap may be used as an early selection criterion for rose rootstocks.An attempt was made to describe correlative inhibition in rose. The apex of the primary shoot is thought tot exert apical dominance over the axillary buds along the shoot, until the terminal flower is visible. As at that time, the auxin export from the apex decreases, the inhibition of the uppermost 2-4 axillary buds is released, resulting in bud break. Later, the young axillary shoots will take over apical dominance and inhibit bud break further down the stem by their auxin production. When the terminal flower buds of the axillary shoots are visible, apical dominance is reduced, leading, following the same reasoning as above, to basal bud break.The effectivity of ZR in side-shoot formation, as was shown in in vitro experiments, combined with the fact that ZR is the major translocation form of cytokinins in the xylem points at a key role for ZR or its immediate metabolite, which might be zeatin, in growth and development of rose plants.