Abstract: Auto white balance plays a key role in a digital camera system, and determines image quality to a large extent. An adaptive auto white balance algorithm for digital camera is proposed. With a new color temperature estimation method based on the correlation between Y, Cb and Cr, the algorithm finds more proper white pixels to calculate the averaged chromatic aberration to improve the precision of the estimated color temperature. The algorithm is realized in the design of a digital camera processor. The test results indicate that the precision of the estimated color temperature is improved dramatically compared with the classical algorithms.