Tilt-table testing is one means of quantifying the static roll stability of highway vehicles. By this technique, a test vehicle is subjected to a physical situation analogous to that experienced in a steady state turn. Although the analogy is not perfect, the simplicity and fidelity of the method make it an attractive means for estimating static rollover threshold.The NHTSA has suggested the tilt-table method as one means of regulating the roll stability properties of light trucks and utility vehicles. One consideration in evaluating the suitability of any test method for regulatory use is repeatability, both within and among testing facilities. As a first step toward evaluating the repeatability of the tilt-table method, an experimental study examining the sensitivity of tilt-table test results to variables associated with methodology and facility was conducted by UMTRI for the Motor Vehicle Manufacturers Association. This paper reports some of the findings of that study.The test program examined the influence of twelve specific variables. The matrix of tests consisted of the baseline condition and thirty-three off-baseline conditions. One hundred and thirteen individual tilt-table tests were conducted on a single test vehicle. In general, the tilt-table method was found to be quite robust in that measurement results appeared to be rather insensitive to most of the variables considered. Surface friction under the low-side tires and trip rail geometry, when applicable, were the primary exceptions to this rule.