Electrical energy can be converted to thermal energy associated with the temperature gradient between the two opposite sides of a thermoelectric module. The hot and cold sides of a thermoelectric module can be used as a heater and cooler, respectively. To study the effectiveness of using the cold side as a cooler for frosting systems, five sets of experimental apparatus with different cooling methods for the hot side are set up. These cooling methods include three active cooling methods based on (a) fan enhanced heat pipe cooling, (b) forced convection water cooling, and (c) forced convection air cooling, and two passive cooling methods based on (d) free convection water cooling, and (e) free convection air cooling. In the experiments, frosting mass and transient temperatures are recorded, and thermal images for frost surfaces are taken. Both the transient and time averaged coefficients of performance (COP) for module, cooler, and frosting are calculated and compared based on the transient temperatures. It is interesting to observe that the transient frosting COP of cooling method (e) decreases to negative values with time, implying that it is not practical for frost growth. Moreover, the comparison results show that the average module COP based on active cooling methods is relatively high compared to passive cooling methods, implying that cooling fluid water is more efficient than air. The ranks of average cooler and frosting COPs for the five cooling methods are in the order of (a) > (b) > (c) similar to (d) > (e).