The real and perceived risks of hydrogen fuel use, particularly in passenger vehicles, will require extensive safety precautions including hydrogen leak detection. Conventional hydrogen gas sensors require electrical wiring and may be too expensive for deployment in multiple locations within a vehicle. In this recently initiated project, we are attempting to develop a reversible, thin-film, chemochromic sensor that can be applied to the end of a polymer optical fiber. The presence of hydrogen gas causes the film to become darker. A light beam transmitted from a central instrument in the vehicle along the sensor fibers will be reflected from the ends of the fiber back to individual light detectors. A decrease in the reflected light signal will indicate the presence and concentration of hydrogen in the vicinity of the fiber sensor. The typical thin film sensor consists of a layer of transparent, amorphous tungsten oxide covered by a very thin reflective layer of palladium. When the sensor is exposed to hydrogen, a portion of the hydrogen is dissociated, diffuses through the palladium and reacts with the tungsten oxide to form a blue insertion compound, H{sub X}WO{sub 3}- When the hydrogen gas is no longer present, the hydrogen will diffuse out of the H{sub X}WO{sub 3} and oxidize at the palladium/air interface, restoring the tungsten oxide film and the light signal to normal. The principle of this detection scheme has already been demonstrated by scientists in Japan. However, the design of the sensor has not been optimized for speed of response nor tested for its hydrogen selectivity in the presence of hydrocarbon gases. The challenge of this project is to modify the basic sensor design to achieve the required rapid response and assure sufficient selectivity to avoid false readings.