Abstract: Objective To establish a new method for PFGE fingerprint analysis, which is easy-to-operate, inexpensive, software-low-dependent and readily-exchangeable.Methods A group of 44 vibrio parahaemolyticus was subtyped according to PulseNet standardized PFGE protocol. The fingerprint was analyzed with new gel-partition bands counting method and BioNumerics clustering separately and then the results by the two methods were compared with type number, cluster number and Simpson's Index of Diversity (DI).Results The 44 isolates could be typed into 32 types and 5 clusters were found by gel-partition bands counting method (DI=97.6%). BioNumerics could type the same isolates into 29 types and 4 clusters were found (DI=95.5%). The difference in the ability of finding clusters between the two methods was not statistically significant.Conclusion Gel-partition bands counting method is based on the bands distribution among finger print and greatly reduces the number of visually observed spectrum, which is easy-to-operate, inexpensive, software-low-dependent and readily-exchangeable.