Most of the spatially referenced data held by the Kenya national examination council (KNEC) are in analogue hard copy format. This necessitates large storage facilities for storing the paper maps, which have low retrieval speeds. Additionally, wear and tear are occasioned during retrieval and handling, and sometimes some of the data is lost. In this form data sharing is difficult and reproduction usually involves high costs per unit. The purpose of this paper is to implement a geographic information system (GIS) which will lower cost per unit, by allowing higher retrieval speeds, smaller storage facilities requirements, while facilitating data sharing. This GIS will perform all the tasks of the current manual system and in addition, provide functionality to aid in the efficient management of the Kenya national examination council data. To accomplish this, existing hardcopy data was digitized and cleaned. New data was collected, processed, analyzed and stored in the form of a geodatabase. This geodatabase stores both the spatially related data and the attribute data. This geodatabase can be used to answer many questions, but for this work, we emphasize the aspect of efficiency in exam distribution. To determine the most efficient routes to follow in the distribution of examinations during the examination period, a geometric network was prepared which was then used to determine the best routes. In this research, a prototype GIS has been developed. Visualization and comparison can be easily performed using the digital maps produced from the implemented system. The GIS database created can be used for purposes of querying and can be revised whenever new information is available. Shortest distances analysis and efficient distribution route determination were performed using spatial analysis and network analysis tools. From the distribution analysis, the service area analysis is demonstrated as giving a more realistic spatial extent of coverage compared to the buffering approach. From these analyses, the services area analysis and buffering approach showed areas of 980.96 and 223.15 Ha being beyond zone 5. Key words: