The author attempts by an account of the very early treatment of a small twin girl to modify Winnicott's theory of transitional phenomena. On the basis of observation of this child, the author concludes that the child plays a more active part in the constitution of areas of transition between his mother and himself than Winnicott believed. The child has a minimum basic defence and anxiety in him, such that, although the object retains its transitional qualities, it sustains the illusion that the infant created it itself to a lesser extent than Winnicott believed.