Lack of mechanistic models to describe petrophysical properties of mobile phases in gas hydrate‐bearing sediments is one of the key challenges in accurately predicting gas production. The major drawback of empirical models that are used to fit a relative permeability curve for mobile phases in gas hydrate‐bearing sediments is that they are based on experimental data, which are limited, and not able to account for hydrate morphology in pore space. This study proposes a relative permeability model that is mechanistic in nature and developed to account capillarity by building on the original nonempirical relative permeability model that assumes negligible capillary pressure. The proposed model implicitly accounts for capillarity with the help of four empirical parameters (two for each mobile phase) that incorporate each mobile phase pressure. It is shown that the proposed model provides an improved match to relative permeability data (derived from pore‐scale simulation of gas hydrate‐bearing sediments) than nonempirical relative permeability by accounting for the effect of capillary pressure. Additionally, unlike fully empirical models that can predict relative permeabilities reliably only at gas hydrate saturations for which experimental data are available, the proposed model only requires fitting the empirical parameters once with experimental data at any single gas hydrate saturation and then it can then be used to predict relative permeability at any gas hydrate saturation. The mechanistic nature of the proposed model allows studying relative permeability of hydrate‐bearing sediments as a function of hydrate morphology and wettability (fluid phase distribution) besides other physical parameters of the model (e.g., porosity, gas, and water residual saturation). Based on the sensitivity analysis of different hydrate morphologies on gas/water relative permeability, it is found that gas relative permeability is sensitive to hydrate morphologies, while water relative permeability shows little dependency on hydrate localization in pore space.