We investigated the hypothesis that the critical heat flux (CHF) occurs when some point on a heated surface reaches a temperature high enough that liquid can no longer maintain contact at that point, resulting in a gradual but continuous increase in the overall surface temperature for most power-controlled systems. This hypothesis unifies the occurrence of the CHF with the quenching of hot surfaces by relating them to the same concept: the ability of a liquid to contact a hot surface, generally defined as some fraction of the liquid’s homogeneous nucleation temperature, depending on the contact angle. The proposed hypothesis about the occurrence of the CHF is investigated through a study of the boiling mechanism of the second transition region of nucleate pool boiling of water on copper. An idealized two-dimensional transient conduction heat transfer model was developed to investigate the heat transfer mechanism. The initial macrolayer thickness on the dry portion of the heater, in the second transition region, was found to be bounded between 0 and 11 μm. The radius of the dry patch varied from 15 to 23 mm (60 and 92 percent of the heater radius, respectively) for initial macrolayer thicknesses of 0 and 11 μm, respectively. The results indicated that the critical liquid-solid contact temperature at the onset of CHF (the surface temperature at the center of the dry patch) must be lower than the homogeneous nucleation temperature of the liquid for the pool boiling of water on a clean horizontal surface. The liquid-solid contact temperature was dependent on the initial dry patch liquid macrolayer thickness, varying from 180°C to 157°C for initial macrolayer thicknesses of 0 and 11 μm, respectively. Independent assessment of these values shows good agreement with extrapolated contact temperature data at the onset of film boiling. This indicates that the mechanism for the occurrence of the CHF could be similar to the mechanism generally accepted for the quenching of the hot surfaces. Further study of this mechanism to understand better the observed trends in other experimental results show qualitative agreement with those results. These include a significant decrease in the radius of the dry patch to 4 mm (16 percent of the heater radius) when the thermal conductivity of the heater was decreased to that corresponding to nickel. When the thickness of a copper heater was decreased from 10 mm (representing an infinitely thick medium) to 0.1 mm, a dry patch radius of 2.25 mm (9 percent of the heater radius) was found to be sufficient for the temperature at the center of the dry patch to reach the critical contact temperature. These comparisons are felt to provide some understanding as to why the second transition region has been observed only in limited cases.