This paper deals with a new blind watermarking technique using two dimensional Walsh coding. The aim of using the 2D Walsh coding is to improve the robustness of the algorithm. The watermark which is a hand written signature was encoded by using 2D Walsh functions then it was embedded in the low frequency coefficients of the discrete cosine transform of the host image. The new algorithm is blind and does not require the original image to extract the watermark and cause minimal distortion to the host image. The robustness of the algorithm was assessed against various Stirmark attacks such as JPEG compression, noise, and some filtering operations. The extent of the improvements is related to the scaling factor.