Compared with traditional chemical analysis methods, reflectance spectroscopy has the advantages of speed, minimal or no sample preparation, non-destruction, and low cost. The present study explored the application of the reflectance spectroscopy within near ultraviolet-visible-near infrared region to predict bio-element compositions in the ornithogenic sediments from the maritime Antarctic. A total of 106 samples were taken from four ornithogenic sediment cores on the Ardley Island of Antarctica, 68 samples were used for building calibration equation, and 38 for prediction of nine bio-elements including P, Ca, Cu, Zn, Se, Sr, Ba, F and S. Three multivariate statistical analysis techniques, including stepwise multiple linear regression (Stepwise-MLR), principal component regression (PCR) and partial least squares regression (PLS) were used to develop mathematical relationships between the spectral data and the chemical reference data. The results showed that the regression models constructed by PCR and PLS models have no significant differences, and obviously supervisor to Stepwise-MLR. The correlations between spectra-predicted and chemically analyzed concentrations of nine bio-elements are statistically significant, and the concentration-versus-depth profiles predicted from reflectance spectra using PLS calibration model are consistent with those from actual chemical analysis. These results demonstrated the feasibility of using reflectance spectroscopy to infer bio-element concentrations in the ornithogenic sediments, and thus it is suggested that the reflectance spectroscopy could provide a rapid and valuable technique to indirectly identify whether the sediments were influenced by penguin droppings in the Antarctic region.