Competitive pressures in the integrated circuit (IC) industry are now requiring short time-to-market product introductions. With the most successful of these products ultimately reverting to commodity status, it is also becoming important to be able to convert from flexible manufacturing within a facility set up for engineering assisted processing to low-cost volume production as experience is gained. Computer-based design for manufacturability (DFM) of the IC process is the only effective way to meet this challenge, but today the unavailability of detailed information about process tool margins is preventing true DFM of process technology from becoming a reality. Statistical metrology is a method consisting of a short-loop process flows and parametric electrical test that can supply the margin performance data of specific process tools needed for DFM. Test structures are designed with large numbers of test sites and with a variety of sensitivities to the process variations. Each structure uses the minimum tool set and process flow necessary for electrically measurable parameters so as to ease the inversion of statistically confounded data. By utilizing technology computer-aided design (TCAD) tools to simulate the short process flow and by assigning statistical distributions to individual steps, a best fit to the data provides useful calibration data of key TCAD parameters. These parameters can then be used to simulate the process margin of a full process and, through DFM, make the process optimally robust for manufacturing.