Ice concentration measurements are important information e.g., for navigation, ice modeling, and climate change research. Here we present an algorithm for estimating ice concentration from C-band SAR data. The resolution of SAR data is significantly higher than the resolution of the current operational ice concentration products based on radiometer data. Our algorithm is based on segment-wise autocorrelation distributions. The algorithm results were compared with two reference data sets: ice concentrations from the Finnish Ice Service (FIS) ice charts, and ice concentrations from the radiometer-based operational ice concentration algorithm of University of Bremen. The new algorithm gives reasonable ice concentration estimates in a high resolution (500 m) for an arbitrary segmentation.