With the popularity of new media, the relationship between the media and the public has changed considerably. A comprehensive and quantitative analysis of the relationship between the public and media can reveal the development of news media in China and help provide some constructive advice to improve their quality. Therefore, we establish vector autoregression (VAR) models between the media and the public on 160 network trending events from 2011 to 2018 based on the Baidu Index. In specific terms, we explore the causal relationship between them with the Granger causality test and analyze the dynamic effects using the impulse response function (IRF) analysis. Our findings suggest that there are satisfactory two-way interactions between the news media and the public in China (especially in the event categories directly related to people’s livelihood, such as safety misadventure, natural disasters, food and drug safety, and entertainment) and that the public plays a leading role for most events. We also find that the news media lags behind users generally, which prompts us to propose the three-relay hypothesis to explain its dissemination mechanism. Also, the impulse responses of attention to negative events are generally more drastic than those to positive events. For hot network events in China, the time span and sample size of our research are large, increasing the results’ reliability.