Although significant progress has been made in the development of robots with serpentine properties, the issues of motion control and adaptation to environmental constraints still require substantial research. This is particularly true for search and rescue applications, where reliable operation in extremely difficult terrain is essential. This paper presents a novel robot design based on the mechanics and neural control of locomotion in Caenorhabditis elegans, a tiny nematode worm. Equipped with an extremely simple yet powerful neurally-inspired decentralized control system, the robot presented here is capable of effective serpentine locomotion. More importantly, it exhibits sensorless path finding, in which obstacles in the environment are overcome, based purely on proprioceptive feedback encoding body shape. Indeed, the robot lacks any form of external sensory capability. The design and implementation of the prototype robot and its control strategy are discussed. In order to validate the control strategy for path finding, experiments and analyses have been performed. The results show that the robot can find its path successfully in the majority of cases. The current limitations have also been discussed. © 2012 IEEE.@en