Double-ended interferometry has several benefits over single-ended gauge block interferometry: there is no need for wringing, which wears surfaces and requires expertise, and there is improved repeatability, since there is no variation due to inconsistent wringing conditions or form errors of the gauge block surfaces. Some disadvantages of double-ended interferometry are that absolute phase change correction is needed for the gauge block and its uncertainty has a double effect on total uncertainty. In addition, elimination of the wavefront error is more complicated than with single-ended interferometry. A simple optical modification that enables double-ended interferometer (DEI) measurements with the MIKES interferometer for long gauge blocks is presented. This modification is applicable to almost any single-ended interferometer (SEI). A procedure for evaluating the wave front correction for different parts of the interferogram of DEI is explained, and a modification and software with capability for nine-point phase stepping is presented. Three independent methods for evaluation of the phase correction were studied. One of them uses integrating sphere for the surface roughness correction and literature values for the phase change due to complex refractive index of material correction. The second evaluates the phase correction from the difference between DEI and SEI results obtained with a quartz platen. The third uses differences-from separate measurements-between the results obtained with quartz or steel auxiliary platens. Only a few gauge blocks per set need testing to obtain phase correction. SEI and DEI results with different phase correction determination methods are presented and evaluated. The uncertainty 2 9 2 .