Waste foundry sand (WFS) is known as the main waste material of foundry industries, and its disposal cost and environmental threats have become one of the major challenges in many countries. To promote the reuse of WFS, it can be utilized as a partial substitute for fine aggregates in concrete technology, which reduces disposal costs, adverse environmental effects, and the need to extract natural aggregates. The aim of this study is to predict the compressive strength of concrete containing WFS (CCWFS) using an innovative hybrid technique i.e. biogeography-based optimization (BBO) assisted by an artificial neural network (ANN) as an affordable alternative to minimize the reliance on experimental activities. To develop the proposed model, a comprehensive database (340 mix designs) including effective parameters on the compressive strength of CCWFS is compiled from the open literature. The architectural structure of the proposed model and its performance evaluation were analyzed by a variety of statistical indicators. The results indicated the potential for reliable performance of the proposed BBO-ANN model in predicting the compressive strength of CCWFS. Finally, to validate the proposed model, a comparison was made with results of five recent studies with different prediction models, in which our findings indicate that the proposed BBO-ANN model offers by far the best performance in predicting the compressive strength of CCWFS compared to the other ones. This study provides new insights into applying the hybrid predictive technique that can help future research to predict properties of green concrete containing waste materials more accurately.