A knowledge-based system, called visualization tool assistant (VISTA), which was developed to assist scientists in the design of scientific data visualization techniques, is described. The system derives its knowledge from several sources which provide information about data characteristics, visualization primitives, and effective visual perception. The design methodology employed by the system is based on a sequence of transformations which decomposes a data set into a set of data partitions, maps this set of partitions to visualization primitives, and combines these primitives into a composite visualization technique design. Although the primary function of the system is to generate an effective visualization technique design for a given data set by using principles of visual perception the system also allows users to interactively modify the design, and renders the resulting image using a variety of rendering algorithms. The current version of the system primarily supports visualization techniques having applicability in earth and space sciences, although it may easily be extended to include other techniques useful in other disciplines such as computational fluid dynamics, finite-element analysis and medical imaging. Senay, Hikmet and Ignatius, Eve Unspecified Center NAS5-30428; RTOP 506-59-11...