The contrastive analysis is made based on the data from automatic weather station (AWS) and conventional station during the parallel observation from January to December 2003 at the Binhai Weather station in Yancheng. The differences of relative humidity between two observing systems are analyzed, and the relationships of relative humidity, ambient temperature, weather condition and wind speed to the relative humidity difference (ΔH) between automatic and conventional stations are discussed. The results show that there are some differences between automatic and manual observations of relative humidity in Binhai County. The bias is within the allowable range of accuracy, about 93% of data biases are between-4% to 4%. The relative humidity is less than 70% when the absolute value of bias is larger than 2%. The annual distribution of ΔH appears Wshaped fluctuation. With the increase of air temperature, the distribution of ΔH shows Vshaped variation tendency. When the air temperature is high, the magnitude of auto weather station humidity is usually smaller than that of manmade observation. When humidity increases or it rains, ΔH tends to be positive. When wind speed increases, ΔH has a decreasing trend, and the relative humidity of the automatic station is larger than that of manual observation.