Analysis of the proportion of cell types in native wood is important for understanding the environmental. stresses including an increasing atmospheric CO, concentration on the structure of wood, especially for the management of plantation forests which will reduce our reliance on natural forests, The conventional method for determining the proportion of cell types is a quantitative microscopy, which is one of the image analyzing systems using a light microscope combined with a microcomputer. However, it is a lengthy multistep procedure. We have examined the feasibility of using Fourier transform Raman spectroscopy for rapid determination of proportion of cell types (fiber, ray parenchyma, vessels, and axial parenchyma) in native wood with using wood meals of two Eucalyptus species, including samples of various ages and colors. By the application of second derivative transformation of Raman spectroscopic data and the partial least-squares regression, we have successfully obtained highly significant correlations between microscopically measured and Raman predicted values for all traits except vessels with correlation coefficients of > 0.9 and 0.8, respectively, in the calibration and in the prediction. This method is valid for all traits since vessels can be calculated by the rest of three traits, and will help to solve the effect of the environmental issues on trees and the supplement of renewable raw materials. (C) 1999Academic Press.