User interest modeling is crucial for personalized news recommendation. Existing personalized news recommendation methods usually take the news data as the minimum interest modeling unit when modeling user’s interests. They ignored the low-level and high-level signals from user’s behaviors. In this paper, we propose a news recommendation method combined with multi-level and multi-interest user interest modeling, named MMRN. In contrast to existing methods, our MMRN model captures user interest at multiple levels, including word-level, news-level and higher-levels, then learns multiple user interests vectors in each level. Extensive experiments on a real-world dataset show our method can effectively improve the recommendation effect.