In the present study, the relationship between the results of ultrasonic testing as a non-destructive inspection with tensile strength and fatigue life of three-sheet spot-welded joints was investigated using a single and dual-objective neural network. Since both destructive tests (tensile and fatigue) cannot be performed on the one specimen, in order to implement the dual-objective neural network that predicts tensile strength and fatigue life simultaneously, a new approach was used that similar ultrasonic results have tensile strength and fatigue life equality. For validation of this approach, the results of the dual-objective neural network were compared with single-objective neural networks (the separate prediction of tensile strength and fatigue life). Firstly, the parameters of the near-optimal neural networks were determined by high repetition and finally, the NSGA II was used to optimize the final structure of neural networks. The NSGA II results indicated that the tensile strength and fatigue life for the dual-objective neural networks have about 6 % and 2 % difference with the respective single-objective neural networks. Also, the two-objective neural network has about 5.5% and 2 % difference in tensile strength and fatigue life, respectively with actual results. The most important achievement of this study states that the new approach of the dual-objective neural network can be used to the inspection of spot-welds in the automotive industry.