This paper considers whether a tax on currency transactions could be expected to raise a significant sum of money for use in the war on poverty. It traces the detailed discussion of Kenen and the subsequent argument of Schmidt, that technical developments would now permit the tax to be levied efficiently by the five authorities who issue currencies in which transactions are settled. It notes the creation of the CLS Bank and the proposal to confine a currency transactions tax to transactions that go through that bank, but argues that this would have dangers. It notes also Spahn's proposal for a geographically limited tax, but argues that this would not be advantageous if the aim is to raise revenue. The final verdict is that a currency transactions tax of 1 basis point would be feasible and could be expected to raise a sum of the order of US$ 20 billion per year.