Knowledge discovery in data sources available on computational grids is a challenging research and development issue. Several grid research activities addressing some facets of this process have already been reported. This chapter introduces the GridMiner framework, developed within a research project at the University of Vienna. The project's goal is to deal with all tasks of the knowledge discovery process on the grid and integrate them in an advanced service-oriented grid application. The GridMiner framework consists of two main components: technologies and tools, and use cases that show how the technologies and tools work together and how they can be used in realistic situations. The innovative architecture of the GridMiner system is based on the Cross-Intdustry Standard Process for Data Mining. GridMiner provides a robust and reliable high-perfomance of grid-enabled applications in terms of e-science and detailed analysis of very large scientific data sets. The interactive cooperation of different services - data integration, data slection, data transformation, data mining, pattern evaluation and knowledge presentation - within the GridMiner architecture is the key to productive e-science analytics.