Rapeseed is one of the most important oil crops in the world, but its oil quality and seed yield are affected by the genetic purity of the varieties. Plant variety identification plays a vital role in maintaining genetic purity leading to improve seed business. Current methods for rapeseed variety identification include visual inspection and modern techniques such as DNA profiling. The former which are based on phenotypic character may be prone to error while the later may be expensive and cannot be performed on-site. NIR spectroscopy offers a rapid and non-destructive approach that could overcome these limitations. This study is aimed to evaluate the potential of portable/handheld NIR to make a supervised classification model for the rapeseed varieties. The seed samples (N=225) of three (03) rapeseed varieties were scanned with handheld SCiO NIR sensor and the average of the three scans were used for classification of varieties. The classification model developed by the combination of different pre-processing and classification algorithms were tested on unknown samples (n=75). It was found that all classifiers exhibited good results except Partial Least Square–Discrimination Analysis (Rc²=0.8). SIMCA classification was tested which correctly identified 96.4% and 93.3% samples from training and test sets respectively followed by Random Forest classifier (F1=0.97) with a success rate of 93.3% on test set. However, Support vector machine (C-SVM type) with a polynomial kernel (3^rd degree) gave accurate results after a combination of Standard Normal Variate (SNV) and first order Savitzky-Golay derivative (polynomial degree of 2) with number of smoothing points (window size) of 5. It classified 100% samples of training set and 97.3% samples of test set into their correct classes. Based on initial evaluation of four classification algorithm, it was found that SVM can be better utilized for varietal classification. This study reveals that handheld NIR can be a reliable and useful tool for rapeseed variety identification, which can benefit both the seed industry as well as the farmers.