Heat waves (HWs) exert severe impacts on ecosystem, social economy, and human lives. Thus, changes in HWs under a warming climate have triggered extensive interests. In this study, the authors developed a new method to identify the HW events in China by double thresholds and further classified them into four categories (i.e., mild, moderate, severe, and extreme HWs) according to their magnitudes by using the daily maximum temperature data from 701 observation stations. On this basis, the spatiotemporal features of HWs with different severities in China from 1961 to 2015 were investigated. The results show that the high HW frequency mainly appears in Jianghuai, South China and western Northwest China. Moreover, the high frequencies of moderate, severe, and extreme HWs occur from June to August and reach the peak in July, while the mild HW frequency is compared from May to September. Since the 1960s, the frequencies of the mild, moderate, severe, and extreme HWs in China have increased significantly with rates of 7.5, 4.3, 1.4, and 1.8 events per year, respectively. The increases are the greatest in July for the moderate, severe, and extreme HWs while comparable during May to September for the mild HW. Besides, an interdecadal change is found to occur in the late 1990s. Compared with the former period (1961–1996), the occurrence of the extreme HWs during the latter period (1997–2015) has increased most significantly in eastern Northwest China and North China, while the frequency of the mild HWs increases most significantly in Jianghuai and South China.