Due to the relatively small amount of in situ data available for the open oceans, remote sensing techniques take an important role in the retrieval of geophysical information, particularly during extreme events. The work presented here aims at the improvement of prediction of cyclone intensity using synthetic aperture radar (SAR) images. A new method to measure the hurricane intensity using SAR images, in combination with a parametric Holland-type model of wind speed, is presented. The algorithm is based on a least square minimization of the difference between the parametric model results and the SAR measurement. The radius of the maximum wind speed, required as input for the minimization procedure, is estimated from the SAR image using wavelet analysis. Information on wind direction is extracted from the SAR image through analysis of image features caused by boundary layer rolls. The root-mean-square error of the suggested method has been validated to be equal to 3.9 m/s. The study is based on a data set of wide-swath SAR images of about 400 km × 400 km coverage, acquired by the European Envisat satellite, over tropical cyclones. As a case study, hurricane Katrina is investigated in detail. A total of five tropical cyclone images will be used to validate the results of the new algorithm.