The conditions for the metalorganic molecular beam epitaxial growth of Hg1−xCdxTe (x = 0.18−0.32) alloys at very low growth temperatures (T ≤150°C) have been optimized by correlating the surface properties and crystalline perfection with the incident Hg flux. A window for growth has been defined for x = 0.18, 0.23, and 0.32. A thermodynamic model has been developed to account for void formation. A neural net model has been used for the first time to model the dependence of void density on the Hg flux and the x-ray rocking curve widths on growth parameters. The combination of these two complementary modeling techniques allows for a flexible process optimization to be carried out with a minimum effort spent in calibration runs.