To account for the positions in which vascular cambia regenerate in wound callus, a gradient induction hypothesis was proposed in 1961 in terms of gradients in 'some factor as yet unknown'. It now seems likely that the gradient is based on morphogen diffusion between source and sink on opposite sides of existing cambia, with morphogen diffusing into the adjoining wound callus. It is specifically proposed that there are two morphogens, auxin diffusing centrifugally and sucrose diffusing centripetally. The cambium then regenerates along a path where the ratio of auxin to sucrose concentration is similar to that at the original cambium, and its orientation (as regards xylem and phloem formation) is determined by the direction of the gradient in this ratio. These proposals are supported by published evidence on auxin and sucrose concentration gradients across the cambium, and on their sources, movements, and known effects on vascular differentiation. Simulations of the proposed positional control system predict patterns of cambial regeneration and orientation corresponding to those observed in four different types of wound and graft.