Use of data assimilation techniques such as optimal interpolation or the Kaiman filter in global chemistry transport models (CTM) is becoming more common. However, owing to high computational requirements, it is often difficult to apply these techniques to multidimensional models containing extensive photochemical schemes. We present a sequential assimilation approach developed for use with general global chemistry transport models. It allows fast assimilation and mapping of satellite observations and provides estimates of analysis errors. The suggested data assimilation scheme evolved from the one described by Levelt et al. . It is a variant of the suboptimal Kaiman filter and is based on ideas described by Menard et al.  and Menard and Chang [200O]. One of the most important features of the developed scheme is its ability to routinely estimate variance of the analysis and to predict variance evolution in the model. The developed technique (or its variants) has been successfully interfaced with a number of different global models and used for assimilation of several types of measurements, including aerosol extinction ratios. Some of these experiments are described by Lamarque et al.  and W. D. Collins et al. (Forecasting aerosols using a chemical transport model with assimilation of satellite aerosol retrievals: Methodology for INDOEX, submitted to Journal of Geophysical Research, 2000, hereinafter referred to as Collins et al., submitted manuscript, 2000). We illustrate the method using assimilation of ozone observations made by the Upper Atmosphere Research Satellite/Microwave Limb Sounder in the three-dimensional chemistry transport model ROSE [Research for Ozone in the Stratosphere and its Evolution; Rose and Brasseur, 1989].