The present invention relates to a neural network-based sharing car intelligent optimization decision method and system. The neural network-based sharing car intelligent optimization decision method comprises the steps of firstly obtaining the set number of training sample pairs, wherein each training sample pair comprises an input training sample and a corresponding output desired value, then establishing a BP neural network, utilizing the training sample pairs to carry out the iterative training on the BP neural network, and finally inputting the to-be-detected input parameters in the trained BP neural network to obtain the corresponding nearest car taking position, a favorite car type, an optimal driving path, the cost-optimal consumption total price and a terminal point car returning position, help people to evaluate the travel parameters comprehensively, and decide the reasonable paths, thereby achieving the purposes of searching the optimal driving experiences and improving the vehicle selection modes, and effectively avoiding the vehicle selection and travel blindness and randomness.