Infrared radiation associated with vapor-liquid phase transition of water is investigated using a suspension of cloud droplets and mid-infrared (IR) (3–5 μm) radiation absorption measurements. Recent measurements and Monte Carlo (MC) modeling performed at 60 °C and 1 atm resulted in an interfacial radiative phase-transition probability of 5 × 10−8 and a corresponding surface absorption efficiency of 3–4%, depending on wavelength. In this paper, the measurements and modeling have been extended to 75 °C in order to examine the effect of temperature on water's liquid-vapor phase-change radiation. It was found that the temperature dependence of the previously proposed phase-change absorption theoretical framework by itself was insufficient to account for observed changes in radiation absorption without a change in cloud droplet number density. Therefore, the results suggest a strong temperature dependence of cloud condensation nuclei (CCN) concentration, i.e., CCN increasing approximately a factor of two from 60 °C to 75 °C at near saturation conditions. The new radiative phase-transition probability is decreased slightly to 3 × 10−8. Theoretical results were also calculated at 50 °C in an effort to understand behavior at conditions closer to atmospheric. The results suggest that accounting for multiple interface interactions within a single droplet at wavelengths in atmospheric windows (where anomalous IR radiation is often reported) will be important. Modeling also suggests that phase-change radiation will be most important at wavelengths of low volumetric absorption, i.e., atmospheric windows such as 3–5 μm and 8–10 μm, and for water droplets smaller than stable cloud droplet sizes (<20 μm diameter), where surface effects become relatively more important. This could include unactivated, hygroscopic aerosol particles (not CCN) that have absorbed water and are undergoing dynamic evaporation and condensation. This mechanism may be partly responsible for water vapor's IR continuum absorption in these atmospheric windows.