AbstractCritical Shape Metrology (CSM), a Critical Dimension Scanning Electron Microscope (CD‐SEM)‐based technique that extracts accurate feature shape information from images obtained during routine in‐line wafer inspection as a means of minimizing measurement bias, is described and explored experimentally. CSM uses intensity profiles from CD‐SEM images of known materials that are compared in real time to profiles contained in an off‐line generated Monte Carlo SEM simulation library. The library of intensity waveforms spans the range of expected geometric feature shapes and material compositions. Extensive comparison of CSM results to accepted reference measurement system measurements using Critical Dimension Atomic Force Microscopes, cross‐sectional SEMs and Focused Ion Beams are made. Agreement of the CSM method to each of these is shown to be better than 1% or within experimental uncertainty and significantly better agreement than traditional CD metrology algorithms. Copyright © 2005 John Wiley & Sons, Ltd.