Abstract: In this study, data from 33 typical coastal stations (station data) during 1981-2012 were used to calculate mean monthly sensible heat fluxes and latent heat fluxes to evaluate two types of reanalysis data, i.e., the ERA-Interim (ERA-I) and the MERRA reanalysis monthly sensible flux and latent flux data. The objective was to provide a reference for the user of ERA-I or MERRA heat flux data. The results indicate that:(1) from the monthly mean sensible heat flux and latent heat flux data, neither the average of the ERA-I data nor the average of the MERRA data deviated from the station data for the whole coast of China. The mean value of MERRA monthly heat flux was closer to the mean value of the station data than ERA-I's. Both the ERA-I and MERRA data were signi ficantly correlated with station data. There were signi ficant differences in variance between the reanalysis data and the station data. The differences of variances between ERA-I data and the station data were less than those between MERRA data and the station data. (2) Analysis of the seasonal changes suggests that both ERA-I data and MERRA data have signi ficant correlations with station data for the whole coast of China. The reanalysis data could re flect the seasonal variations in sensible heat flux and latent heat flux.There are some differences noted. For example, in both sets of reanalysis data, sensible heat fluxes from February to July were greater than the station data; the amplitude of seasonal change in latent heat flux from the reanalysis data was less than the station data. The lowest seasonal changes occurred in the MERRA data. (3) The analysis of inter-annual variability from 1981 to 2012 showed that both the ERA-I heat flux data and the station heat flux data had declining trends for the whole coast of China. There was a signi ficant correlation between ERA-I data and the station data. The MERRA sensible heat flux data had a slightly declining trend. The MERRA latent heat flux data series showed a slightly increasing trend and showed different trends compared with the station latent heat fluxes. The correlation was not significant between MERRA heat flux and station heat flux. The heat flux data from MERRA did not represent the inter-annual change features for the whole China coast.