An initial experiment has been conducted to verify geostationary-satellite-derived cloud motion wind estimates with in situ aircraft wind velocity measurements. The experiment was conducted during December 1972 over the Caribbean Sea with the following aircraft: the National Center for Atmospheric Research (NCAR) Buffalo and Sabreliner, and NASA and U.S. Air Force RB-57F's. The Buffalo and Sabreliner were used for in situ wind velocity measurements and cloud height determinations, and the RB-57F's for precision aerial photogrammetry. Case histories of ½ to 2 h were obtained for 3–10 km diameter cumulus cloud systems on 6 days. Also, one cirrus cloud case was obtained. In most cases the clouds were discrete enough that both the cloud motion and the ambient wind could be measured with the same aircraft Inertial Navigation System (INS). Since the INS drift error is the same for both the cloud motion and wind measurements, the drift error drops out of the relative motion determinations. The magnitude of the vector difference between the cloud motion and the ambient wind at the cloud base averaged 1.2 m s−1. The wind vector at higher levels in the cloud layer differed by about 3 to 5 m s−1 from the cloud motion vector.