In this research, take the seventh legislative elections for target, It’s first time two ticket voting for congress in Taiwan and it’s a good chance to observe split -vote havior. The unit of analysis are 358 towns in this country. Thronging conventional regression and spatial analysis model to understand the factors that affect the split- voting. Not only the split voting rate as the dependent variable, but also the Pan-Blue and pan-green camp split voting rate .the study found that the most important factor affecting split vote are the third force candidates, showing the charm of candidates for parliamentary elections has a very important influence. The factor that influence blue camp split vote rate are agricultural population and ethnic group. The factor that influence green camp split vote rate are urbanization and ethnic group. The spatial analysis, prove that the split-vote rate would be affected by the proximity effect, the split –vote behavior in adjacent area will be similar.