By extracting the layer-charges inYBa2Cu3Oxas functions of dopant oxygen contentxfrom neutron diffraction data (using the bond-valence-sum method), we show that the charge-transfer hypothesis of high-temperature cuprate-plane superconductivity fails to describe the data. A similar failure occurs forLa2−&bgr;Sr&bgr;CuO4as a function of Sr dopant concentration, and was implicitly predicted (before the charge-transfer hypothesis was enunciated) by self-consistent local-density-approximation calculations of R. V. Kasowski, M.-H. Tsai, J. D. Dow, and M. T. Czyzyk [Physica C162, 1349 (1989)]. Both failures can be traced to difficulties the hypothesis has with (i) the rules of chemistry which require electronegative dopants such as oxygen to attract electrons, not holes, and (ii) the law of conservation of (neutral) charge in each unit cell. ©1997 American Institute of Physics.