Rotating or stationary targets were presented simultaneously with a random-dot mask which was itself either rotating or stationary. Observers were required to judge both the form of the target and its motion; these target attributes were manipulated orthogonally in the experiment. The results support the hypothesis that under these conditions figural information is encoded independently from motion information. On a trial-by-trial basis within conditions of target and mask motion combinations, performance in the two tasks was independent. Across conditions, different effects were observed in the two dependent measures. Form judgments were generally more accurate for rotating than for stationary targets, particularly when the mask was stationary. Motion judgments showed frequent intrusions from the motion of the mask, so that most errors occurred when target and mask were dissimilar in terms of their motion. The results show differential interference of the random-dot mask on the encoding of the two independent stimulus attributes, and support the possibility of separate processing of figural and motion information.