In contrast to text documents, research data, with its graphic visualizations, places different demands on indexing, searchability and presentation in the information retrieval process. The aim of the VisInfo project is the development and prototypical implementation of innovative approaches for interactive, graphical access to research data, to present it in the information retrieval process and make it searchable in the best way possible. In the project, we have studied and developed further data analysis processes as well as visual search systems, with their prototypical implementation being evaluated for research data from earth and environmental sciences. In this talk we will present the general challenges of a visual search approach in research data and the VisInfo prototype as outcome of the project.