Recently Wang et al. [J. Chem. Phys. 117, 10184 (2002)] reported fragilities for various molecular liquids, as determined by calorimetry. The technique is based on measuring the change in fictive temperature with cooling rate. We point out that the same method was previously applied to polymers [Robertson, Santangelo, and Roland, J. Non-Cryst. Solids 275, 153 (2000)], yielding good agreement with fragility values obtained by mechanical or dielectric spectroscopy. (C) 2003 American Institute of Physics.