The thermal degradation kinetics of pyridoxine hydrochloride were examined using a dehydrated model food system designed to simulate a ready-to-eat breakfast cereal. This study was carried out to provide information useful in estimating the thermal losses of pyridoxine hydrochloride in the processing of breakfast cereals and other low moisture dehydrated food systems. Portions of the model system which were fortified with pyridoxine hydrochloride were toasted in a conduction oven at 155°, 170°, 185°, and 200°C for a minimum of six heating times at each temperature. Pyridoxine (PN) content was determined in the heat treated model systems by high performance liquid chromatography (HPLC). For each heat treatment, the loss of pyridoxine could be described by a first order kinetics model. The first order rate constants for 155°, 170°, 185°, and 200°C were 1.74 × 10−2 min−1, 5.22 × 10−2 min−1, 16.88 × 10−2 min−1, and 48.95 × 10−2 min−1, respectively. The calculated Arrhenius activation energy was 29.8 Kcal/mol. In comparing the HPLC method to the standard microbiological assay, the HPLC assay gave lower PN values for the toasted model system. To explain this discrepancy, HPLC fractions were collected and analyzed by the microbiological assay. No significant vitamin B6 activity was found in any fraction other than that containing the PN peak. It is possible that the milder extraction procedure used in the HPLC assay is less capable of recovering forms of PN which may become bound during the toasting process. These potentially bound forms may or may not be biologically available.