This paper discusses findings obtained by measuring power consumption using two different types of power measurement in CNC machines. Readings taken from CNC machines (using PicoLog Current Data Logger) generally have differences in comparison to the actual power (using a dynamometer). Therefore, this paper analyzed the output differences between the two measurement methods and discussed how significant the differences are as well as the relative trends of both results when the same parameters are applied. The experiment used an extruded mild steel on a CNC lathe where cutting speed, feed rate and depth of cut were used as input parameters. Power consumed (converted from energy measured) from both means of measurement were set as the output. The data were analysed and the best combination parameters for machining was proposed in minimising power consumption using the Taguchi Method.