Motor vehicle powered by regular gasoline is one of major sources of pollutants for local and global environment. The current study developed and validated a new fuel-economy optimization system (FEOS), which receives input from vehicle variables and environment variables (e.g., headway spacing) as input, mathematically computes the optimal acceleration/deceleration value with Lagrange multipliers method, and sends the optimal values to drivers via a human-machine interface (HMI) or automatic control systems of autonomous vehicles. FEOS can be used in both free-flow and car-following traffic conditions. An experimental study was conducted to evaluate FEOS. It was found that without sacrificing driver safety, drivers with the aid of FEOS consumed significant less fuel than those without FEOS in all acceleration conditions (22-31% overall gas saving) and the majority of deceleration conditions (12-26% overall gas saving). Compared to relative expensive vehicle engineering system design and improvement, FEOS provides a feasible way to minimize fuel consumptions considering human factors. Applications of the optimal model in the design of both HMI for vehicles with human drivers and autonomous vehicles were discussed.