Abstract.A survey of the global linearization problem is presented. Known results are divided into two groups: results for general affine nonlinear systems and for bilinear systems. In the latter case stronger results are available. A comparision of various linearizing transformations is performed. Numerous illustrative examples are included.1. Introduction. Exact linearization of nonlinear control systems is probably the most intensively studied and best understood area of the differential geometric approach to the nonlinear control. It presents a natural and challenging task: to find suitable compensations that make a given nonlinear system to behave in a linear fashion, or at least closely to it. This problem deserves a lot of attention: its positive solution directly extends the applicability of the linear methods to a more general nonlinear class of systems. As a consequence, beginning with the pioneering works of Krener [24] and Brockett [3], up to the present time there has been increasing interest in the various modifications of the exact linearization problem. Survey of this field can be found in papers [16,31] or in books [22,29].In spite of being aware that all these surveys do not cover many important subfields like time-scaling transformations [33] or dynamic feedback [7,8], this paper does not aim to give an updated survey of the overall linearization area. It will be concentrated only on the global aspects of the exact linearization since a systematic and self-contained exposition on this topic is still missing.The number of results on global linearization is rather modest in comparision with the above mentioned linearization boom. The main reason for this is that