Abstract: In the safety analysis of nuclear reactor accidents, mass and heat transfer between vapor and liquid phases are critical factors influencing the accuracy of safety analysis codes. Developing an interfacial heat transfer package tailored to two-fluid system codes holds significant engineering application potential and research value. Through a comprehensive investigation and evaluation of interfacial heat transfer models used in internationally recognized reactor system analysis codes and published literature, a new interfacial heat transfer package was proposed. The package included three key components: flow regime map, interfacial area models, and interfacial heat transfer correlations. The flow regime map encompassed all relevant flow regimes, including bubbly flow, slug flow, annular-mist flow, inverted slug flow, and inverted annular flow. For interfacial area determination, different calculation methods were employed based on the specific characteristics of each flow regime, and suitable heat transfer correlations were selected to ensure accurate computation. To verify the accuracy of the new developed interfacial heat transfer package, it was integrated with the RELAP code and extensively tested against a wide range of steady-state and transient tests. The steady-state results demonstrate strong alignment between predictions from the new code and experimental data. In the Christensen test, the axial void fraction predictions from the new code show an overall relative error of less than 10%, validating its accuracy. In the OSV test, the code effectively captures the void fraction trends. For the ORNL test, wall temperature predictions closely match experimental values, while axial void fraction and vapor temperature predictions are slightly higher but remained within acceptable error margins, demonstrating overall reliability. Additionally, in the MIT pressurizer test, the pressure drop predicted by the new code occurs more rapidly than the experimental measurements. However, the overall trend is consistent, and the results remain satisfactory. In the FLECHT SEASET steam cooling tests, the results show that the new code provides more accurate predictions of interfacial heat transfer processes under most conditions compared to RELAP, with comparable accuracy in other conditions. For the FLECHT SEASET vaporization tests, the new code and RELAP exhibit similar accuracy at lower axial positions, but at higher positions, the accuracy of RELAP is significantly lower than that of the new code. In the Bennett test, RELAP significantly overestimates the interfacial heat transfer coefficient, leading to overly optimistic wall temperature predictions, whereas the new code optimized these predictions, offering more accurate results. In conclusion, the interfacial heat transfer package developed in this study can accurately simulate interfacial heat transfer processes. Its accuracy surpasses that of the RELAP interfacial heat transfer model, particularly in flow regimes occurring after CHF (critical heat flux).