The price surveys from the 2005 International Comparison Program (ICP) imply substantially lower levels of GDP at purchasing power parity (PPP) for many developing countries than prior estimates. While some observers have questioned the data, this paper argues that the pattern of changes in PPPs between ICP rounds makes economic sense. Consistently with the original Balassa–Samuelson model, more rapidly growing economies experienced steeper increases in their PPPs relative to official exchange rates. This effect was even stronger for poor countries. Taking account of this effect would reduce the need for such large data revisions when new ICP data become available.