Many advanced manufacturing processes employ scanning electron microscopes (SEM) for on-line critical measurements for process and quality control. This is the third of a series of papers discussing various causes of measurement uncertainty in scanned particle beam instruments, and some of the solutions researched and developed at NIST. Scanned particle beam instruments especially the scanning electron microscope have gone through tremendous evolution to become indispensable tools for many and diverse scientifi c and industrial applications. These improvements have signifi cantly enhanced their performance and made them far easier to operate. But, ease of operation has also fostered operator complacency. In addition, the user-friendliness has reduced the need for extensive operator training. Unfortunately, this has led to the concept that the SEM is just another expensive digital camera or another peripheral device connected to a computer and that all of the issues related to obtaining quality data have been solved. Hence, a person (or company) using these instruments may be lulled into thinking that all of the potential pitfalls have been fully eliminated and they believe everything they see on the micrograph is always correct. But, as described in this and the earlier presentations this may not be the case. The fi rst paper in this series discussed some of the issues related to signal generation in the SEM, including instrument calibration, electron beam-sample interactions and the need for physics-based modelling to understand the actual image formation mechanisms to properly interpret SEM images. The second paper, discussed another major issue confronting the microscopist: specimen contamination and methods of contamination elimination. This third paper, discusses vibration and drift and some useful solutions.