A procedure is described for obtaining road meter roughness index measurements directly from profilometer transducer outputs, eliminating the intermediate step of double integrating the profilometer accelerometer signals. (Double integration of accelerometer signals is the major data processing task of inertial profilometers.) Simulation of a linear car model is performed with two separate sets of equations of motion. The first set of equations, which accepts height sensor data, is identical in form to the equations used in simulations where the pavement profile is the input data. The second set is a transformation of the first, arranged to accept accelerometer data. Vehicle responses to the two components of pavement profile are added together and the roughness index is then calculated in the normal way. In the case of road meters that use relative displacement (or velocity) to calculate the roughness index, the two sets of equations can be combined into one set with the input changed to a combination of height sensor and accelerometer signals. The procedure is tested using sinusoidal and random signals that correspond to various elevation profile curves. A comparison is also made with results from the normal procedure of using the elevation profile for the simulation. Results from the testing confirm that the proposed procedure is valid and that explicit computation of the elevation profile is not necessary when the roughness index is measured with a profilometer.