The relationship between garment design parameters, wearer morphology, posture, fabric properties and ease allowance values is very complex and uncertain due to the presence existence of various human factors related to designer and consumer perception. The aim of this work is to set up a series of mathematical models to characterize and determine the relationship between these parameters. For this purpose, fuzzy modelling techniques were proposed to extract IF-THEN rules by learning from the experimental data. First, the general modelling principle and related concepts were outlined. The different samples collected to obtain the learning data were described in detail. The samples were then characterized by instrumental measurements and sensory evaluations to form the input and output data for modelling. Based on the obtained input and output learning data, it became possible to determine the modelling procedure to extract relevant fuzzy rules. The computation then helps in determining the 3D ease allowance values from the desired human perception on fit, comfort and other design parameters. For a new sample and consumer morphology, the relevant degree of input data with respect to all the fuzzy rules was also calculated to find the most appropriate ease allowance value. The developed mathematical model followed by experiments also provides advantages in controlling and adjusting patterns according to perceptions of the designer and consumer. The concept of determining 3D ease allowance could be further integrated into the 3D clothing CAD approach to obtain a suitable garment surface on specific virtual human model environments to customize any garments in the fashion industry.