The invention discloses an intelligent diagnosis method and system for health state of a power distribution network. The intelligent diagnosis method comprises the steps: calculating the value of eachkey feature quantity in a key feature quantity set based on the predetermined key feature quantity set and the obtained current power distribution network data; substituting the value of each key characteristic quantity into a pre-constructed power distribution network health index prediction model to obtain a health index; and obtaining the health state of the power distribution network based onthe health index. For the intelligent diagnosis method, the proposed key feature quantities can solve the problems of complex characteristic quantity and high redundancy of power distribution networkhealth evaluation, and can provide a theoretical basis for power distribution network health index calculation; and due the proposal of the power distribution network health index prediction model, deep mining can be performed on large-scale and wide-range power distribution equipment and complex and changeable power distribution network data, thereby effectively preventing health deterioration of the power distribution network, and preventing faults or power failure accidents.