The application of flow boiling in microchannels in copper cooling elements for very high heat flux dissipation from microprocessor chips is one of the promising technologies to replace air cooling and water cooling of these units, particularly in mainframes and servers. Recently, the authors have proposed a new theoretical model to predict the critical heat flux (CHF) in microchannels, and it is used here to perform a parametric study to investigate the effects of fluid, saturation temperature, mass flux, inlet subcooling, microchannel diameter, and heated length on CHF for this application. The parametric study shows that CHF is increased by: (i) decreasing channel length, (ii) lowering saturation temperature, (iii) increasing mass flux, (iv) increasing inlet subcooling, and (v) increasing microchannel diameter. The best coolant is water, but water is not feasible for the present application because of its very low saturation pressure at 30-40C. Of the other four fluids simulated, their order of merit from best to worst is as follows: R-245fa, R-134a, R-236fa, and FC-72. FC-72, however, has a low saturation pressure (in fact, it would operate under vacuum at the saturation temperatures of 30-40C envisioned here) and is not a candidate fluid for the flow boiling coolant here. Furthermore, the authors have also recently proposed a diabatic flow map for microchannels based on their database for R-134a and R-245fa in 0.5- and 0.8-mm channels. The new CHF model has been incorporated into their map here to predict the transition from annular flow to dry-out, which is a critical design limitation for microprocessor coolers. Importantly, this map then provides the feasible operating range of such coolers with flow boiling as the cooling process, in terms of mass flux and maximum vapor quality at the outlet to avoid CHF.