The advances in sensor network, linked data, and service-oriented computing has indicated a trend of information technology, i.e.
, toward an open, flexible, and distributed architecture. However, the existing information technologies show a lack of effective sharing, aggregation, and cooperation services to handle the sensors, data, and processing resources to fulfill user’s complicated tasks in near real-time. This paper presents a service-orientated architecture for proactive geospatial information services (PGIS), which integrates the sensors, data, processing, and human services. PGIS is designed to organize, aggregate, and co-operate services by composing small scale services into service chains to meet the complicated user requirements. It is a platform to provide real-time or near real-time data collection, storage, and processing capabilities. It is a flexible, reusable, and scalable system to share and interoperate geospatial data, information, and services. The developed PGIS framework has been implemented and preliminary experiments have been performed to verify its performance. The results show that the basic functions such as task analysis, managing sensors for data acquisition, service composition, service chain construction and execution are validated, and the important properties of PGIS, including interoperability, flexibility, and reusability, are achieved.