Social responsibility investment (SRI) has attracted worldwide attention for its potential in promoting investment sustainability and stability. We developed a three-step framework by incorporating environmental, social, and governance (ESG) performance into portfolio optimization. In comparison to studies using weighted ESG rating scores, we constructed a data envelopment analysis (DEA) model with quadratic and cubic terms to enhance the evidence of two or more aspects, as well as the interaction between the environmental, social, and governance attributes. We then combined the ESG scores with financial in-dicators to select assets based on a cross-efficiency analysis. The portfolio optimization model incorporating ESG scores with selected assets was constructed to obtain a social responsibility investment strategy. To illustrate the effectiveness of the proposed approach, we applied it in the United States industrial stock market from 2005 to 2017. The empirical results show that the obtained SRI portfolio may be superior to traditional investment strategies in many aspects and may simultaneously achieve the consistency of investment and social values.