Commercial development of natural polyisoprene from polyisoprene-producing plants requires detailed knowledge on how to select high-polyisoprene-content lines and establish agronomic cultivation methods for achieving maximum polyisoprene yield. This development can be facilitated by a high-throughput quantification method for natural polyisoprene. In this paper, we describe the Fourier transform near infrared spectroscopy (FT-NIR) technique coupled with a partial least squares (PLS) regression model to quantify natural polyisoprene in Eucommia ulmoides leaves. PLS regression models are discussed with respect to linearity, root-mean-square error of estimation (RMSEE), and root-mean-square error of prediction (RMSEP). The best PLS regression model was obtained with second derivative NIR spectra in the region between 4000-6000 cm(-1) (R2Y, 0.95; RMSEE, 0.25; RMSEP, 0.37). This is the first report to employ FT-NIR analysis for high throughput and solvent-free quantification of natural polyisoprene in leaves.