Problems recently discovered in the Kerley method of estimating age at death from cortical microstructure are discussed. Kerley's original data have been re-analyzed to produce new regression equations and to document the original field size.