Abstract: In the research of optimizing the laying density of carbon fiber in remediation black-odorous water，with the help of the statistical software ( SIMCA-P) ，this paper explored the difference of the single indicator optimization method and principal component analysis method by using four factors ( COD，TP，NH3－N and TN) as analysis variables． The results showed that: Single indicator optimization method could confirm the best laying density of the four indicators respectively，while principal component analysis method could comprehensively analyze the information of four indicators and confirm the best carbon fiber laying density as 40 cm2 /L． The study could provide reference for the integrated use of carbon fiber．