During the cutting processes, the tools are usually adjusted to their initial setting periodically in order to produce high quality products. The model of the residuals of the tool wear process in traditional machine tools is fitted using Glejser test method which can test the increasing or decreasing heteroskedasticity of the process. Thus, the process capability index is developed for tool wear process. The Bernstein probability density function is applied to describe the usual characteristics of tool life and the tool wear process. Based on the adjustment cycle which satisfies the required minimum process capability index of customers, integrating process capability and cost, a tool replacement model which determines the optimal initial tool setting, tool adjustment cycle and tool replacement time is presented with regard to quality loss, replacement cost, penalty cost for possible tool failure, adjustment cost and salvage value. A possible application of the model is given that provides a method of off-line tool management in a production environment.