One of the problems in the two-phase flow is investigating the type of flow pattern that occurs in a mini channel without experimenting first. Computational Fluid Dynamics (CFD) is one way that can be used to predict pressure drop, flow pattern, void fraction, other parameters in fluid flow through a channel of a specific size. This research used the CFD simulations to predict the pressure gradient in a two-phase flow of air-sodium chloride 0.9% and glucose in a horizontal capillary tube with a diameter of 1.6 mm, and length of 70 mm. This research aims to obtain the simulation results of 2D and 3D pressure gradients to validate the experimental data. The simulation was carried out with a variation of the superficial gas velocity 0.1 - 8.3 m / s against the superficial velocity of the liquid constant 0.207 m / s. Simulations were carried out using the Volume of Fluid (VOF) model with the "ANSYS Fluent 2020 R1" software. In 2D and 3D simulations, the gradient values of pressure gradients tend to increase as JG increases. The pressure gradient fluctuates linearly at first, then exponentially from mid-to-end. The findings demonstrate that simulations may properly model physical processes like pattern creation and phase interactions. However, due to several physical elements considered in 3D simulation but missed in 2D simulation, the findings of the 2D and 3D simulations were dramatically different when looking at the pressure gradient values.