Most of the numerical studies of the cavitation process available were simulated based on the Rayleigh-Plesset equation, which has a presumed value for the number of bubble nuclei in the domain. However, in the computational domain for most cases, the bubble nuclei density varies, depending on the hydrodynamics and different bubble nucleation rates. Therefore, the modeling of the bubble nucleation rate based on the number of molecules required to form a critical cluster, which is affected by the hydrodynamics in the domain, is one of the challenges that must be overcome for bubble nuclei density prediction. In this work, an enhanced Classical Nucleation Theory (CNT) model was written as in-house code for integration in Computational Fluid Dynamics (CFD) to estimate the water vapor bubble nucleation rate across a quasi-stable cavitating flow nozzle. The enhanced CNT model included microscopic factors that specifically address water vapor bubble nucleation. Because these factors vary in the radial direction, this work was extended into the 2D domain to estimate the bubble nucleation rate in both the axial and the radial directions. The bubble nucleation rate was estimated through various numbers of molecules (depending on the hydrodynamics in the domain) required to form a critical cluster. The Population Balance Model was used to estimate the bubble nuclei density based on the bubble nucleation rate. The simulated results indicated that the bubble nucleation rate was highest at the throat of the nozzle, especially near the wall of the nozzle. Thus, a higher bubble nuclei density was found near the wall after the throat of the nozzle. © 2015 The Author(s). Published by Taylor & Francis.