In this study, applicability of the thermal network analysis (TNA) technique in predicting the frost growth on cold fins was tested. Based on the frost formation model by Lee et al. (1997) and the analogy between the heat and mass transfer, time variations of the thickness and the density of the frost layer were predicted. At the same time, the temperature distributions of the frost layer, air stream and the fin were obtained for each time step. Then, the results were compared with the solutions by full CFD (Computational Fluid Dynamics) analysis as well as with the experimental results for two- and three-dimensional cases that have been reported by others previously. Although the results of the TNA show somewhat lower accuracy compared with those of the CFD analysis, the technique is much simpler and cost-effective in quick estimation of the frost-layer growth.