This paper proposes a multi-criteria optimization technique using the mathematical models developed by the response surface methodology (RSM) for the target responses combined with desirability indices for the determining the optimum cutting parameters in end milling of AISI D2 hardened steels. Different responses may require different targets either being maximized or minimized. Simultaneous achievement of the optimized (maximum or minimum) values of all the responses is very unlikely. In machining operations tool life and volume metal removed are targeted to be maximized whereas the machined surface roughness need to be at minimum level. Models showing the combined effect of the three control factors such as cutting speed, feed, and depth of cut are developed. However, a particular combination of parameter levels appears to be optimum for a particular response but not for all. Thus adoption of the method of consecutive searches with higher desirability values is found to be appropriate. In this study the desirability index reaches to a maximum value of 0.889 after five consecutive solution searching. At this stage, the optimum values of machining parameters - cutting speed, depth of cut and feed were determined as 44.27 m/min, 0.61 mm, 0.065 mm/tooth respectively. Under this set condition of machining operations a surface roughness of 0.348 μm and volume material removal of 7.45 cm3 were the best results compared to the rest four set conditions. However, the tool life would be required to compromise slightly from the optimum value.