This paper aims to revise the critical flow criterion and flashing inception of the analytical critical flow model and to further explore the effect of upstream conditions on critical mass fluxes. The flashing inception of the model is considered to be affected by the Reynolds number and the Jakob number. Model predictions show strong similarities with test results compared with other models, with average errors of less than 10.1% for simulated slits and 15.8% for natural cracks. Upstream parameters of the liquid, such as the stagnation pressure, subcooling, and aspect ratio, directly influence the rate of depressurization, the thermodynamic nonequilibrium constant, and the friction pressure drop, respectively, which determine the critical mass flux. This research provides the theoretical basis for accurate assessment of critical crack, providing safeguards for monitoring nuclear reactor leaks.