Carbon monoxide, even at low ppm levels, may dramatically reduce the performance of a fuel cell stack if not remediated. Remediation generally requires parasitic bleeding of a small fraction (< 5 %) of air into the fuel stream to promote oxidation of the CO. For enhancement of efficiency, this must be controlled using a real-time feedback of CO level in the feed-stream. In this paper, a novel data-driven pattern identification method is applied for robust online sensing of continuously changing CO content in a reformed fuel stream. The pattern identification algorithms have been built upon the underlying principles of Symbolic Dynamics, Information Theory and Automata Theory. The sensitivity of the CO sensor can be tailored for a particular application. Experiments were performed on a 5 cm2 fuel cell to demonstrate the feasibility of the Symbolic Dynamic technique for measurement of CO. A similar approach is now being used to develop on line sensors for a variety of other important fuel cell phenomena, such as flooding and catalyst degradation.