In this paper, a parallel ray-casting volume rendering algorithm based on adaptive sampling is presented for visualizing TB-scale time-varying scientific data. The algorithm samples a data field adaptively according to its inner variation, and thus sets sampling points only in important positions. In order to integrate adaptive sampling into the parallel rendering framework, an efficient method is proposed to handle the resulting unstructured sampling data. The experiments demonstrate that the proposed algorithm can be used to effectively render inner data features in high quality.