To understand the evolutionary characterization of HA1 of H1N1 influenza virus HA gene circulaing from 1981 to 2005 in China, viral RNAs of 370 H1N1 strains were extracted and transcribed into cDNA by reverse transcriptase and amplified by PCR. The products of PCR were sequenced. The sequences were analyzed through biometic software. The results showed that all the four antigenic sites were mutated, bigger change occurred on the Sb and Ca sites; the 130 loop of receptor binding sites(RBS) of HA1 amino acid deleted at the 134th site in 1991 firstly, then the number of the deleted strains were increasing, since 2000, all the strains had deleted at the 134th site, and simultaneously, the amino acid at 137th site was substituted by S for T. The change of HA1 glycosylation sites was found and 7 sites kept stable from 2000 to 2005. The H1N1 strains of the same year almost clustered in the same group on the phylogenetic tree and were irrelevant to virus isolated time and area. There appeared two groups of 2005 H1N1 virus strains that differed in time of virus isolation.