This work attempts to address challenges of using magnetic sensors for target detection, localization and tracking with a wireless sensor network (WSN). A WSN comprised of magnetic sensors was constructed to investigate the modeling, detection, and localization of ferrous targets. The system was established as a centralized tree-based wireless network with a PC acting as the fusion center. A heavy cylindrical iron bar was used as a test target and modeled as a magnetic dipole. The magnetic signal models used are experimentally validated and the problem of variation in magnetic sensor sensitivity is addressed by including sensitivity estimates in the validation process. Maximum likelihood and least-squares techniques were applied to detect and localize the test target. Experimental results demonstrating performance are presented and practical considerations of utilizing magnetic sensors in a WSN addressed.