The Raman spectrums of hemoglobin from 20 normal persons and 21 esophageal cancer patients were measured and analyzed in the present paper, the results show obvious differences between the Raman spectrums of esophageal cancer patients and normal persons. Compared with normal persons, there are more iron ions in low spin state and less in high spin state in the hemoglobin of esophageal cancer patients. It indicates that some iron ions in high spin state turned to low spin state because of cancerization, and this experimental result coincides with the fact that hemolysis is more likely to occur in cancer patients. Principal component analysis(PCA) was employed to get a three-dimensional scatter plot of PC scores for the health and cancer groups, and it can be learned that they are distributed in separate areas. By using the method of discriminate analysis, it was found that the diagnostic algorithm separates the two groups with sensitivity of 90.5% and diagnostic specificity of 95%, and the overall diagnostic accuracy was 92.7%.The results from this exploratory study demonstrate that Raman spectroscopy combined with multivariate analysis would be an effective method for early diagnosis of esophageal cancer.