An investigation was conducted t o determine the acceleration response of a light airplane to roughness of unpaved runways. landing gear of the airplane utilized cantilever spring-type struts for the main and tail wheels. attitude at constant speeds over a 1,000-foot portion of two grass run- ways and a smooth concrete taxiway. The data from the concrete taxiway was used as a reference. Elevation profiles of the runways and taxiway were measured. The The airplane was taxied in the three-point One runway was considered to be as smooth and the other as rough as runways that would likely be encountered at an operational airport. The rough runway, which had a mean deviation of 0.254 foot, was nearly four times rougher than the other which had a mean deviation of 0.066 foot. Average peak-amplitude incremental center-of-gravity acceleration varied from about 0.07g at 5 miles per hour to 0.23g at 40 miles per hour on the rough runway and from 0.06g at 15 miles per hour to 0.14g at 40 miles per hour on the smooth runway. On the concrete taxiway, maximum response was &out 0.048 at 40 to 45 miles per hour. Mean- square acceleration response to the runways increased approximately linearly with speed up to about 25 or 30 miles per hour but increased more rapidly than the linear rate at higher speeds. time spent above given values of acceleration and airplane acceleration power spectra was also obtained. The percentage of Transfer functions computed from the airplane acceleration response to the two runways were in relatively good agreement at 20 miles per hour. At 40 miles per hour, the apparent lack of satisfactory agreement may have resulted from a nonlinear system and/or from a low level of statistical reliability.