Analysis of biochemical systems requires reliable and self-consistent databases of thermodynamic properties for biochemical reactions. Here a database of thermodynamic properties for the reactions of glycolysis and the tricarboxylic acid cycle is developed from measured equilibrium data. Species-level free energies of formation are estimated on the basis of comparing thermodynamic model predictions for reaction-level equilibrium constants to previously reported data obtained under different experimental conditions. Matching model predictions to the data involves applying state corrections for ionic strength, pH, and metal ion binding for each input experimental biochemical measurement. By archiving all of the raw data, documenting all model assumptions and calculations, and making the computer package and data available, this work provides a framework for extension and refinement by adding to the underlying raw experimental data in the database and/or refining the underlying model assumptions. Thus the resulting database is a refinement of preexisting databases of thermodynamics in terms of reliability, self-consistency, transparency, and extensibility.