Increased environmental protection legislation forces manufacturers to employ environmentally conscious design and manufacturing methods. In addition, customer preferences for energy efficient and environmentally sustainable products influence manufacturers design strategies. In order to influence customer buying behaviors for environmentally friendly products, manufacturers need to understand customer preferences first. Manufacturers can make optimal design decisions based on inference on customers' decision making models. It is recognized that consumers are heterogeneous in their response to different attributes for any given type of product or service. In this paper, we proposed a framework for incorporating heterogeneous customer preferences with Dirichlet Process mixture model for product positioning in environmentally conscious design. The uncertainty about the functional form of the customer preference distribution can be expressed by using a nonparametric prior, in which the number of clusters grows without bound as the amount of data grows. An automobile design case study is used here to demonstrate the proposed approach.