Zinc deficit is one of the most important reasons for botanical disease. It can occur in many forms, for example, as rice blast, bacterial stripe, bacterial blight, and false smut. All these diseases will greatly lead to lower yields during rice harvest. In these paper, we show that Raman spectroscopy provides a single, fast, and sensitive method for detecting zinc deficit, especially during early forms of zinc deficiency. When a plant lacks zinc, the carotenoid content will increase, and many other components' content will decrease, such as starch, proteins, sugars, and amide. All of these changes will be brought out in the Raman spectrum of the rice leaves. In this study, we scanned the healthy leaves, the leaves with early zinc deficiency, and the leaves with zinc deficiency by confocal Raman spectrometer to get the Raman spectra within the wavenumber 200-1,200 cm −1 . First, the ensemble empirical mode decomposition algorithm was used to remove the baseline and the noise in the Raman spectra.Then principal components selected by the successive projections algorithm (PC1, PC2, PC3, and PC6) and characteristic parameters of the spectra S 2 S 1 (the area of 690-780 cm −1 S 1 is divided by the area of 1,080-1,200 cm −1 S 2 ) selected by one-way variance of analysis and related with the content of carotenoid, starch, carbohydrate, soluble sugar, and so forth were the inputs of the least squares support vector machine model. The model is able to determine from a variety of samples whether the sample has zinc deficiency, early zinc deficiency, or if it is healthy, with a 100% accuracy. These results illustrate that early forms of zinc deficit can be detected before the plants show any physical symptoms. This is groundbreaking news for the future development of zinc deficiency monitoring equipment, as well as for other types of deficiency monitoring equipment in general.