Specification for development of the body pattern in the amphibian embryo has usually been thought of as a prolonged process, initiated from an ooplasmic localisation of some kind in what will become the dorsal-anterior midline. The evidence has been interpreted as suggesting that this initial localisation is centred in what will become anterior endoderm, but gives rise by an inductive process in early blastula stages to an overlying organising centre which eventually controls the genesis of mesodermal pattern. Neurectodermal development (especially, the position and pattern of the central nervous system) is seen as controlled considerably later, by inductive signals from submigrating mesoderm at gastrulation. Current work tends to confirm that this sequence of inductive influences can occur at least in experimental situations. It also suggests, however, that in the normal development of the rather small egg of Xenopus, genesis of positional cues that specify the body pattern contributions within the more vegetal material (mes-endoderm) is a rather rapid, widespread and direct consequence of events occurring in the interval between fertilisation and cleavage. Possible molecular bases of early nuclear responses to position within egg material, and the more problematic nature of the positional system itself, are discussed.