Abstract: Motion estimation is the key part of video encoder. It helps remove temporal redundancies in image sequences. But most motion estimation algorithms neglect to take the advantage of the strong correlations within the motion fields. The search windows stay the same through the image sequences and the estimation needs heavy computation. To reduce this computational complexity several algorithms have been proposed. One of the popular method is logarithmic search. This method reduces the complexity but it neglects the high correlation of motion field. Also the number of search points increase as the search range increases. In this proposal we propose a fast motion estimation method that uses "Context Adaptive Search (CAS)" windows. It uses strong spatial correlation of the motion field and the median predictor for motion vector coding. The variable size search window centered by the median values reduces the complexity for motion estimation greatly but also decreases the entropy of motion vectors for motion vector encoding. The simulation results show that the proposed method outperforms the logarithmic search in terms of compression efficiency up to 8% with only one-third search points. The overall PSNR value is also slightly better.