In this study, an attempt is made for the offline prediction of ash and the vitrinite content of froth fraction using the image analysis technique. Average Gray Value (AGV) was used as the key parameter in this study. Time-wise froth images were captured during the experiment and processed in a computer, embedded with Matlab 7.0 software. A good correlation was obtained between AGV and ash content (R-2 0.8309) and also with vitrinite maceral content in concentrate (R-2 0.7166). Low gray value indicates dark froth with a low ash content of coal particles with high vitrinite maceral content, whereas high gray value represents bright froth with more water and high ash clay minerals in the froth.