Life cycle assessment (LCA) is a fundamental tool used in eco-design. However, it can be costly and resource intensive. We take steps toward the automation of the inventory and impact analyses stages of LCA via the proposal and development of a case-based reasoning (CBR) procedure to estimate the ecological effects of a product. The case memory in CBR, which contains representations and ecological effects of known products, is organized using an extension of the function-behavior-structure (FBS) representation for products. The extension includes ecological characteristics and values. We develop similarity metrics to measure the distance between cases in the case memory and the new product. The $k$ -medoids algorithm is used to cluster the case memory, our metrics enable cluster retrieval and case selection, and multiple linear regression analysis is employed for adaptation. Using a database of 100 fans, we test the accuracy of the proposed approach on a cross flow fan not in the database. The method gives ecological effect estimates within 3% of the true values when there are similar fans in the retrieved cluster and about 7% when the retrieved cluster does not contain similar fans.