This paper reviews experimental and computational methods used for simulation of ice accretion on aircraft flying through icing conditions. Such methods were recently reviewed by the AGARD FDP Working Group 20 and the present paper represents a revised and updated version of parts of the Working Group report. To provide essential background, it begins with a brief physical description of the ice accretion process. Experimental simulations must respect certain similarity requirements or scaling laws if they are to be valid; these requirements are discussed in some detail and in the framework of this discussion, physical phenomena are considered in more detail as well. Techniques and ground-based facilities for experimental simulation of ice-accretion phenomena are then reviewed, followed by a review of techniques and facilities used for flight testing in support of aircraft design and certification for flight in icing conditions. Available instruments for required measurements such as droplet size distribution and liquid water content and for inflight ice detection are briefly described. Computational simulation is becoming increasingly important in aircraft icing work; computational methods are used to simulate ice accretion both with and without iceprotection systems in operation. Computational approaches are outlined and current capabilities are evaluated. Conclusions emerging from the review include the following: rime icing is reasonably well understood and can be adequately simulated for most practical purposes using either experimental or computational methods; some of the physical phenomena known to be important in glaze icing are only poorly understood and there is considerable uncertainty regarding whether or not certain other phenomena are important; consequently, much additional research is required before reduced-scale experimental simulations or computational simulations of glaze icing will be sufficiently accurate and reliable for most practical purposes. Research recommendations are put forward.