Wireless sensor networks represent a new generation of distributed embedded systems with a broad range of real-time applications. Examples include medical care, emergency response, wild fire monitoring, and border surveillance. Such systems must achieve real-time performance under severe resource constraints and environmental dynamics. Real-time wireless sensor networks challenge many classical approaches to real-time computing. The special issue consists of three high-quality research papers selected from twenty three submissions. Their research contributions address a diverse set of new real-time issues in wireless sensor networks spanning hardware platforms, network protocols, and real-time scheduling. Mangharam, Rowe, and Rajkumar present FireFly, an integrated hardware and software platform for real-time streaming over wireless sensor networks. FireFly employs a cross-layer architecture comprised of hardware-support for global time synchronization, a collision-free TDMA link layer, and transmission scheduling algorithms tailored for real-time streaming. The effectiveness of FireFly was demonstrated in a 42-node network that achieved sub-20 μs synchronization accuracy for people-tracking and voice communication in a coal mine. Moser, Brunelli, Thiele, and Benini investigate real-time processor scheduling for wireless sensor networks. Many real-time wireless sensor networks must achieve real-time performance over extremely long lifetimes. While energy harvesting has shown promise as an enabling technology for long-running wireless sensor networks,