In this study, temperature sensor signals collected from a wireless temperature sensor were used to estimate MRR in Copper CMP (Cu-CMP). A set of Cu-CMP experiments were conducted on a Buehler Automet 250 polishing machine following L8 Taguchi design of experiments. Material removal and temperature rise (?T) in copper polishing samples was measured during experiments. Regression models based on process parameters (load, relative velocity, and slurry concentration), pad wear factor and temperature rise were developed and compared. Temperature rise values observed during Cu-CMP experiments measured by temperature sensor with a sampling rate of 4Hz were used for estimation of MRR. The predictability of MRR through a regression model comprising of the process parameters only, was low (R2 = 51.7%). The predictability of MRR increased (R2 = 73.5%) after including temperature rise rate as a predictor variable in the regression model. A regression model having ratio of MRR and ?T as the response variable and process parameters, pad wear factor, and their two way interactions as the predictor variables showed further increase in the predictability of MRR (R2 = 82.1%). A regression model having ratio of ?T and MRR as the response variable and process conditions and pad wear factor as the predictor variables improved the accuracy of estimation of MRR (R2 = 87.7%). The improvement in the predictability of MRR is likely because the model accounts for the effect of load, relative velocity, slurry concentration, and pad wear on the slope of relation between ?T and MRR.