Medical imaging devices and systems must be calibrated to ensure uniformity and reliability of test results. A standard reference material (SRM) or "phantom", as it is known in the medical imaging community, is used to replicate fundamental characteristics of tissue and/or the material for which the imaging device is intended for use. SRMs can be readily deployed to sites where devices are being used, negating the need to physically relocate instruments for calibration or performance testing. The SRM produced in this work is appropriate for calibrating depth-resolving 3D optical systems such as optical coherent tomography (OCT). The SRM consists of three layers of polydimethylsiloxane (PDMS) films on a gridded glass slide. A thin, clear layer of PDMS is sandwiched between two thicker scattering layers of PDMS. The scattering layers contain titanium dioxide (TiO2) as scattering particles to allow easy identification of the thicknesses of the three layers as a dimensional calibration of the axial resolution of an optical coherence tomographic imaging device. The concentration of TiO2 scattering particles in PDMS can be adjusted to control its scattering coefficient to mimic that of biological tissue which was independently measured by a broadband integrating sphere system at NIST for these concentrations. The thickness of each layer was measured by NIST's spectral domain optical coherence tomography (SD-OCT) instrument to produce data similar to that obtained using clinical devices. The axial dimensions of each region of interest of the SRM was determined from the pixelated, 3D tomographs acquired with an index of refraction (n) = 1, then converted from pixels to µm using a NIST-traceable height standard to calibrate the NIST SD-OCT and then corrected by the index of refraction of PDMS as measured by spectroscopic ellipsometry at the appropriate OCT wavelength. For the users of the SRMs, we have developed an algorithm which can be applied to a 3D tomograph obtained from any OCT to detect the interfacial planes between the three layers of the SRM. The algorithm reports the local layer thicknesses for each pixel across the entire lateral dimension of the tomographic data cube. The user can then compare the output of the algorithm generated by their OCT tomograph to the values from NIST to determine the calibration of their instrument. For each SRM, we report a mean thickness and standard deviation for each of the three-layer thicknesses across 9 regions of interest as defined by a target grid registered in the SRM.