This paper presents the operational scheme of the National Satellite Meteorological Center (NSMC) of the China Meteorological Administration (CMA) to derive atmospheric motion vectors. The NSMC scheme is compared with a method developed at the European Organization for the Exploitation of Meteorological Satellites (EUMETSAT) in preparation for Meteosat Second Generation. Both schemes employ similar basic principles in terms of feature tracking and height assignment, however there are also some important differences. Furthermore, the EUMETSAT scheme assigns quality indicators to each wind vector at the end of the processing chain, whereas the NMSC scheme has inbuilt quality checking at different processing steps allowing for reinstatement of winds rejected by a first quality check. The evaluation of the performance is gained from two periods: a week in January and a week in July 1999. European Centre for Medium-Range Weather Forecast analyses and radiosonde data are used as independent data for evaluation of the two schemes. It is shown that correlating infrared image data with water vapor data before height adjustment, as performed in the NSMC scheme, has a great potential to better distinguish high and low cloud and to provide high-density wind fields. The utilization of radiative transfer calculations for the estimation of the height of thin clouds in the EUMETSAT scheme is shown to be imperative for good quality wind fields. Finally, the feature of the EUMETSAT scheme to assign quality indicators improves the utility of the wind vectors for use in numerical weather prediction models. It is suggested that a combination of the different features of both schemes potentially provide highly increased spatial density in the wind field with improved quality.