Abstract: An increase in the frequency of heat wave accompanies the global climatic change, and causes the continuous occurrence of grouped heatstroke. In order to reduce the impact of heat wave on human health, an index method for heatstroke meteorological forecast is established based on full research about the method of heatstroke both at home and abroad. The methods on heatstroke weather forecasting include index method, statistical analysis method and weather classification method. The method of statistical analysis is the most widely used now, but needs long-time series observation data. The statistical model is based on correlative analysis between case data and meteorological factors. The daily surface observation data of 358 cities from 1996 to 2005 are used, including daily maximum air temperature and daily average relative humidity. Daily numbers of heatstroke sufferers in Wuhan from June to August in 1994 and 1995 are adopted too. Then the correlation between meteorological factors and heatstroke cases is analyzed, and the heat index is chosen as meteorological index to grade with the method of quartile division considering the climatic background of different regions. Four grades of heatstroke are determined by the daily maximum air temperature and the duration of different high temperature weather condition. The correlations between grades of heatstroke forecast and daily numbers of heatstroke sufferers in Wuhan during July to August of 1994 and 1995 are analyzed, and the grades are also compared with daily maximum temperature. Moreover, days of each grade of heatstroke are statistical analyzed from 1996 to 2005. The results suggest that the grades reflect the influence of weather conditions upon heatstroke reasonably. The method on grading prediction of heatstroke can be applied to the actual operation. Many domestic provincial meteorological bureaus have started heatstroke forecasting research and service these days. However, it is hard to form a national uniform standard due to the vast territory and regional differences of climatic background. The method of quartile division takes different climatic background into consideration, so the standard of heatstroke forecast is graded all over the country. The error is reduced through this method to a certain extent and it provides technique foundation for developing national operational heatstroke forecast method.