Edible marine algae, often referred to as seaweeds, have health benefits and nutritional value. Origin discrimination of seaweeds is essential for quality assurance and traceability. As a rapid, easy, and economical method, near-infrared spectroscopy (NIR) was used in this study to investigate its ability to identify Sargassum fusiforme according to geographical origin, cultivar, and production method. A potential particle swarm optimization-support vector machine (PSO-SVM) identification model was constructed and exhibited its superiority in the origin determination of S. fusiforme. Results showed that the correct recognition rates were 90.00% for geographical origin, 100.00% for cultivar, and 100.00% for production method, when the PSO-SVM models were validated using the test sets. Overall, the results proved the potential of NIR spectroscopy combined with PSO-SVM models as a fast analytical method to trace the origin of S. fusiforme in terms of geographical origin, cultivar, and production method. Display omitted •This work presents a novel approach for the discrimination of Sargassum fusiforme.•NIR spectroscopy was used to identify Sargassum fusiforme from different origins.•PSO-SVM model provided satisfactory discrimination performance.