A hand-held spectrometer can be used to evaluate oil palm (Elaeis guineensis Jacq.) leaf nutrient contents without being destructive. This study aims to develop regression equations and analyze the performance of the prediction models for Nitrogen, Phosphate, and Potassium leaf nutrient contents. The dependent variable in this study was the result of the analysis of nutrient contents in frond number 17 which was carried out in the laboratory, while the independent variable was the leaf reflectance value scanned with a hand-held spectrometer. The Normalized Difference approach is used to create a vegetation index from the combination of reflectance values at two wavelengths. Vegetation index with the highest correlation value to the nutrient content of leaves, is used to make a prediction model for leaf nutrients using the Simple Linear Regression. The regression equations formed to predict the contents of nutrients N, P, and K have high R2. The RMSE values of the predicted contents of N, P, and K nutrients, respectively were 0.21, 0.01, and 0.13; and correctness values of those nutrients respectively were 93.29%, 95.5%, and 88.81%. Keywords: Hand-held spectrometer, Oil palm, Prediction, Leaf nutrients contents.