Previous studies of bank location in Taipei area have neglected using spatial data to analyze. If a study does not use spatial data in analyzing bank location, it would be insufficient and unpersuasive. This study integrated census data and spatial data in Taipei City to analyze the spatial relationship between census data and bank trade area. Through GIS and association rule, one of data mining methods, common characteristics, e.g. spatial and non-spatial factors, of bank locations would be discovered. The result showed that 221 rules (situation) related to bank location were discovered under the condition of supporting level equal and greater than 19% and confidence interval equal or greater than 90%, Moreover these rules would be used build a spatial decision support system for decision maker to evaluate potential locations of bank in Taipei City.