In conventional atmospheric models, isotope exchange between liquid and gas phases is usually assumed to be in equilibrium, and the highly kinetic phase transformation processes inferred in clouds are yet to be fully investigated. In this study, a two-moment microphysical scheme in the NCAR Weather Research and Forecasting (WRF) model was modified to allow kinetic calculation of isotope fractionation due to various cloud microphysical phase-change processes. A case of moving cold front is selected for quantifying the effect of different factors controlling isotopic composition, including water vapor sources, atmospheric transport, phase transition pathways of water in clouds, and kinetic versus equilibrium mass transfer. A base-run simulation was able to reproduce the ~ 50 ‰ decrease in δD that observed during the frontal passage. Sensitivity tests suggest that all the above factors contributed significantly to the variations in isotope composition. The thermal equilibrium assumption commonly used in earlier studies may cause an overestimate of mean vapor-phase δD by 11 ‰, and the maximum difference can be more than 20 ‰. Without microphysical fractionation, the δD in water vapor can be off by about 25 ‰. Also, using initial vertical distribution and lower boundary conditions of water isotopes from satellite data are critical to successful isotope simulations, without which the δD in water vapor can be off by about 34 and 28 ‰, respectively.