The National Institute of Standards and Technology (NIST) has a multifaceted program in atomic force microscope (AFM) dimensional metrology. Three major instruments are being used for traceable measurements. The first is a custom in-house metrology AFM, called the calibrated AFM (C-AFM), the second is the first generation of commercially available critical dimension AFM (CD-AFM), and the third is a current generation CD-AFM at SEMATECH-for which NIST has established the calibration and uncertainties. All of these instruments have useful applications in photomask metrology.Linewidth reference metrology is an important application of CD-AFM. We have performed a preliminary comparison of linewidths measured by CD-AFM and by electrical resistance metrology on a binary mask. For the ten selected test structures with on-mask linewidths between 350 nm and 600 nm, most of the observed differences were less than 5 nm, and all of them were less than 10 nm. The offsets were often within the estimated uncertainties of the AFM measurements, without accounting for the effect of linewidth roughness or the uncertainties of electrical measurements. The most recent release of the NIST photomask standard-which is Standard Reference Material (SRM) 2059-was also supported by CD-AFM reference measurements. We review the recent advances in AFM linewidth metrology that will reduce the uncertainty of AFM measurements on this and future generations of the NIST photomask standard.The NIST C-AFM has displacement metrology for all three axes traceable to the 633 nm wavelength of the iodinestabilized He-Ne laser. One of the important applications of the C-AFM is step height metrology, which has some relevance to phase shift calibration. In the current generation of the system, the approximate level of relative standard uncertainty for step height measurements at the 100 nm scale is 0.1 %. We discuss the monitor history of a 290 nm step height, originally measured on the C-AFM with a 1.9 nm (k = 2) expanded uncertainty, and describe advances that bring the step height uncertainty of recent measurements to an estimated 0.6 nm (k = 2). Based on this work, we expect to be able to reduce the topographic component of phase uncertainty in alternating aperture phase shift masks (AAPSM) by a factor of three compared to current calibrations based on earlier generation step height references.