The output of renewable energy is strongly uncertain and random, and the distribution of voltage and reactive power in regional power grids is changed with the access to large-scale renewable energy. In order to quantitatively evaluate the influence of renewable energy access on voltage and reactive power operation, a novel combinational evaluation method of voltage and reactive power in regional power grids containing renewable energy is proposed. Firstly, the actual operation data of renewable energy and load demand are clustered based on the K-means algorithm, and several typical scenarios are divided. Then, the entropy weight method (EWM) and the analytic hierarchy process (AHP) are combined to evaluate the voltage qualified rate, voltage fluctuation, power factor qualified rate and reactive power reserve in typical scenarios. Besides, the evaluation results are used as the training samples for back-propagation (BP) neural networks. The proposed combinational evaluation method can calculate the weight coefficient of the indexes adaptively with the change of samples, which simplifies the calculation process of the indexes' weight. At last, the case simulation of an actual regional power grid is provided, and the historical data of one year is taken as the sample for training, evaluating and analyzing. And finally, the effectiveness of the proposed method is verified based on the comparison with the existing method. The evaluated results could provide reference and guidance to the operation analysis and planning of renewable energy.