Abstract: Even though China, like any other country, is vulnerable to the adverse effects of climate change, agriculture is a vital industry in the country. In the context of global warming and frequent extreme weather and climate events, agricultural production in China has been affected by the increasingly severe meteorological disasters. It is very important to study the spatio-temporal patterns of agro-meteorological disasters in China in order to mitigate disaster risks and reduce disaster losses. In this paper, the spatial and temporal changes and the typical patterns of main meteorological disasters (i.e., drought, flood, low-temperature and hailstorm disasters) affecting agriculture in China from 1978 to 2013 were analyzed using Empirical Orthogonal Function (EOF) method. The results showed that drought (starting in 2000), hailstorm (beginning in 2001) and low temperature (starting in 2008) all decreased in trend. However, there was no obvious trend in flooding. The EOFs of drought that caused 10% and 30% yield losses were similar, and the losses in the North were obviously higher than those in the South. The EOFs of flood that caused 10% and 30% yield loss were quite different. Flood area with 30% yield loss was mainly distributed in the Yangtze River Basin and the Northeast. However, flood area with 10% yield loss was widely distributed throughout the country. The low temperature disaster in the northern area was more severe than in the southern area. Hailstorm in the western area and the northern area was much serious. Droughts and floods did not affect the northeastern plain throughout the year, but caused serious yield losses. Low temperature disaster persistently affected most areas of China, but with little serious losses. Hailstorm caused persistent and serious losses. Combined with statistical and EOFs analyses, we found that the spatial and temporal patterns and the dynamics of the four disasters with 10% and 30% yield loss were not consistent in disaster degree, affected areas and disaster duration.