In recent years the design of modular products has become the focus of significant research in the area of design theory and methodology. This phenomenon is due to the power of modularity to achieve certain product objectives. Product modularity has many advantages but it can cause reliability problem which relates to simplicity, clarity, and unity of product components. This paper uses fuzzy cluster identification for classifying the functions into several modules and entropy for measuring reliability. We present the relationship between product modularity and reliability according to the number of modules and components.