In this paper, we address the issue of soil roughness characterisation in the case of agricultural fields having different tillage (roughness) states, that is ploughed, harrowed and rolled surface. By using a laser profiler, the surface height variation trace of three types of tillage soil profiles are obtained. Using this dataset the range of RMS height, autocorrelation length and the fractal characteristic values associated with each agricultural roughness state is estimated, The result show that fractal dimension turned out to be a suitable index for distinguish the different tillage states. But there is no one-to-one correspondence relationship. A soil roughness index was found to be appropriate for the quantification of soil surface roughness.