In 1987, a cooperative project between the U.S. Geological Survey and the Institute of Geophysical and Geochemical Exploration was initiated to evaluate the origin of the Hatu gold anomaly. The anomaly is located in the Hatu mining district in the northwest corner of Xinjiang-Uygur Autonomous Region in northwest China. The climate is semiarid to arid and wind erosion predominates. A regional soil survey of the Hatu district, based on samples collected on a 200 by 500 m grid and composited prior to chemical analysis to a density of one sample per square km, delineated a series of south-southeast-trending Au anomalies. Anomalous Au values range from 5 ppb to more than 700 ppb. The Hatu anomaly, the most prominent of these anomalies, is more than 30 km long and about 5 km wide. The mining town of Hatu and the economic gold deposits of Qiqu 1 and Qiqu 2 are at the northern end of this anomaly. The axis of the Hatu anomaly cuts across mapped structure and stratigraphy in the district, but is parallel to the prevailing wind direction. This observation led to the hypothesis that the Hatu anomaly is the result of acolian dispersion of gold from the vicinity of Qiqu 1 and Qiqu 2. The alternative interpretation, that the anomalies reflected additional primary gold occurrences, was not consistent with existing information on the known occurrences and the geology.