An improved suspension-firing reactor has been constructed and a new experimental procedure developed that is able to assess the relative performances of coals and coal blends. The method measures the total combustion time of the samples by measuring the exit gas analysis from the reactor on a continuous basis. The new equipment and procedure overcome some deficiencies that were identified in an earlier suspension-firing method that was based on the extent of burnout of the sample at the end of an arrested combustion test. The presence of a synergistic effect on blending is indicated by a difference between the actual performance of the blend and that predicted by additivity from the performance of the individual coals and the composition of the blend. Three sets of binary blends and their parent coals, which have also been used in utility pf furnaces, have been tested in the suspension-firing reactor. The results have been compared with the performance at the commercial scale, as indicated by the C content of the electrostatic precipitator fly ashes. In all three cases, the performance as predicted by the suspension-firing reactor was consistent with that observed at the commercial scale. This suggests that this relatively straightforward small scale test may provide a useful route for the assessment of coals prior to purchase and for the optimization of blend performance.