Texture is employed on the pad’s surface to improve the performance behaviour of thrust pad bearings. This paper investigates the performance parameters (pressure, load carrying capacity and coefficient of friction) of a rectangular dimpled pad thrust bearing. The effect of dimples’ attributes, such as depth, texture area density, circumferential extent and radial extent, has been explored. The Reynolds equation is solved to compute the performance behaviours for few cases of dimple parameters varying within the given range, yielding the output. An artificial neural network model is trained to predict the performance parameters. It has been investigated that a textured pad having rectangular dimples having a depth of 20 µm, area density of 0.5625, circumferential and radial extent in the range of 0.6–0.8 obtained higher load carrying capacity and lower friction coefficient in comparison to other cases considered in the investigation. The simulated model shows good accuracy and predicts the performance parameters with less than 2% error. Overall, the research showcases the potential of artificial neural networks in predicting bearing performance behaviours.