There is an ever increasing demand on part quality and tighter tolerances for machining of components in high volume manufacturing. A major source of problem in the machine tools is the thermally induced error due to thermal gradients and uneven heating and expansion of various machine components. Current practice of manufacturing precision parts involves periodic gaging of parts, whereby, production is interrupted and manual compensating offsets are input to the controller. Also, additional production costs are introduced due to requirement of initial warm up cycles without cutting parts and utilization of chillers for temperature controlled coolants. In this paper, a methodology is described for automatic compensation for thermal error by means of components/locations temperature profile and calculated error between the tool tip and the workpiece. Global empirical models are developed for predicting the errors in the three axes of the machine using least-squares curve-fitting techniques. During the actual cutting process, the model calculates the errors at the tool tip based on temperature readings. A digital signal processor (DSP) based real time error compensation controller, developed for axes position manipulation, compensates for the errors. The effectiveness of the methodology and the hardware is demonstrated through application for actual cutting tests in a machining center.