A semianalytical algorithm for the retrieval of ice cloud properties from satellite data is presented. The new method is based on the semianalytical cloud retrieval algorithm and uses solutions of the asymptotic radiative transfer theory applicable for optically thick media. Therefore the new method is much less computer time expensive than the commonly used lookup table approaches. Usually, the cloud optical thickness and cloud effective droplet radius are reported for water and ice clouds even though both parameters are dependent on the actual crystal shape assumed in the retrieval procedures. Thus the authors propose to use the reduced optical thickness (ROT) and the particle absorption length (PAL) for the characterization of ice clouds. This implies that no a priori or climatological estimates of the particle shape/size distribution are necessary and increases the comparability of different cloud retrieval algorithms, which are built on many different distribution functions. If still necessary, the retrieved ROT and PAL can easily be transferred to values of the optical thickness and the cloud effective droplet radius by assuming any of those size distribution functions. The developed technique has been applied to data from the NASA EOS Terra Moderate Resolution Imaging Spectroradiometer sensor. The scene shows Hurricane Jeanne just before its landfall near the coast of Florida in September 2004. Both the reduced cloud optical thickness and the particle absorption length have been derived for the eye wall region.