In situ airborne gravity data at altitudes of 11, 6.3, and 1.7 km over a smooth area of Alabama are used to assess gravity accuracy and errors in upward and downward continuations. Analysis of the Alabama free-air anomaly gravity data at crossover points at the three altitudes suggests 1 - 2 mgal accuracy for the dataset. Gravity data at each altitude are then expanded into local 3D Fourier series, to prepare for continuation. This Fourier representation results in continuation errors at few-mgal level in Alabama, even in the extreme case of downward continuation from 11 km to sea level. The result in Alabama inspires an airborne gravity survey over the rough, inaccessible terrain of Tibet. Similar investigations as in Alabama are made in Tibet using EGM08-derived airborne gravity data at flight altitudes of 10, 5, and 0 km. Bouguer anomalies at the 10-km altitude preserve the major tectonic features of Tibet. Downward continuation errors increase with terrain roughness, but the survey can enhance local tectonic features. This study highlights the value of a future Tibetan airborne gravity survey and points out the expected gravity accuracy and spatial resolution from this survey.