Progress in generating robust differentiation protocols for efficient and scalable production of defined cell lineages from human embryonic stem cells (hESc) has been slow. Amongst the obstacles to be addressed are those inherent to standard hESc culture and differentiation practices including the use of feeder cells, serum and animal-derived matrices. These components are biologically complex, undefined and highly variable between batches, inhibiting the development of consistently reproducible protocols (1,2).