In order to evaluate the autonomous combat effectiveness of ground-attack UAV， taking actual combat as a background closely, an autonomous combat effectiveness evaluation method is proposed based on the improved ADC model. Firstly, on the basis of the traditional ADC model, the battlefield environment factors and human interference factors are introduced, and the BP neural network is optimized based on genetic algorithm to reconstruct the traditional combat capability evaluation model, and an autonomous combat effectiveness evaluation model based on ADC-BP is constructed. And then, based on the characteristics of UAV combat mission, the key capability indicators affecting the effectiveness are summarized, and an evaluation index system suitable for the combat process is constructed. Finally, taking suppression of enemy air defense mission performed by a ground-attack UAV as an example, the rationality and practicability of the ADC-BP evaluation model are verified, and this provides a new idea for the evaluation of the autonomous combat effectiveness of ground-attack UAV in the future.