This paper presents a neural network (NN) combined with multiattribute decision making (MADM) approach for determining the design combination of design elements that match a given eco-product value (EPV) and product image. An experimental study identifies 7 office design elements and 27 representative office chairs as experimental samples for developing NN models. A morphological analysis is used to extract 7 form elements from sample office chairs. Fifteen EPV attributes are identified and categorized into aesthetic, functional, and environmental dimensions. A best-performing NN model is chosen to examine the complex relationship between 7 form elements and 15 EPV attributes as well as 6 product images. With the best NN model, an office chair design database is built consisting of 960 different combinations of design elements, together with their associated EPV and product image values. The application of the database provides office chair designers with the best combination of design elements for matching the aesthetic, functional, and environment-friendly attributes as well as specific product image word pairs, thus facilitating the eco-product form deign process.