In the recent years, pervaporation is widely investigated for removal of thiophene compounds from fluidized catalytic cracker (FCC) gasoline. Commercialization of this process depends on the development of suitable membrane modules. The earlier studies were performed on flat sheet membranes. Hollow fiber modules seem to be the most suitable module for this process on account of high packing density for handling the large amount gasoline-stream in the refineries. In this study, a mathematical model is derived for determining the performance of axial flow type hollow fiber module. The experiment could not be performed on hollow fiber module due to unavailability of commercial modules. However, tubular modules are commercially available. Tubular and hollow fiber modules have similar geometry, and their mathematical models are also the same. Therefore, for the model validation, experiments are carried out on a commercially available tubular module for removal of thiophene from n-heptane/thiophene model gasoline. Furthermore, the experimental results from the literature for the ethanol/water separation are also used for model validation. The mathematical model is successfully validated in both cases. Since the performance of membrane modules also depends on different dimensional parameters related to module geometry, the performances of the hollow fiber module are determined by varying these parameters. Simulations show that lower membrane area, lower length/diameter ratio of module, higher module porosity and lower fiber radius leads to improvement in the module performances. Furthermore, the performance of module is compared for both co-current and counter-current flow arrangements, which shows that the values of average fluxes are higher for counter-current flow arrangement. These results suggest that for given conditions, counter-current hollow fiber module is a better option than the co-current hollow fiber module. (C) 2016 Elsevier B.V. All rights reserved.