Abstract Over 100 hair samples were collected from the Xiamen area of south China, including 72 healthy people and 34 cancer patients, for the study of distinguishing cancer patients from others; another 64 hair samples were collected from patients with liver, stomach, lung, and miscellaneous cancers and were used for the discrimination analysis of different types of cancers, The samples were analyzed with ICP-AES for the simultaneous determination of several elements and with GFAAS for Se content after wet digestion. The average values of the element concentrations in the hair samples are presented. The results obtained were then analyzed with principal component analysis and partial least-squares methods, to distinguish cancer patients from others, and with an orthaganolization recurrence method to discriminate different types of cancers. The cancer patients could be completely distinguished from healthy people; while the misdiscrimination rates for the analysis of liver, stomach, and lung cancers are 15.8, 14.3, and 19.0%, respectively.