We describe two different, double-sided interferometer designs for measuring material stability. Both designs are balanced interferometers where the only optical path difference is the sample and the reference beams are located within the interferometer. One interferometer is a double-pass design, whereas the other is a single-pass system. Based on a tolerancing analysis, the single-pass system is less susceptible to initial component misalignment and motions during experiments. This single-pass interferometer was tested with an $86\mathrm{nm}$ thin-film silver sample for both short-term repeatability and long-term stability. In 66 repeatability tests of $30\min$ each, the mean measured drift rate was less than $1\mathrm{pm}/\mathrm{h}$ rms. In two long-term tests ( $>9\mathrm{h}$ ), the mean drift rate was less than $1.1\mathrm{pm}/\mathrm{h}$ , which shows good agreement between the short- and long-term measurements. In these experiments, the mean measured length change was $2\mathrm{nm}$ rms.