Seaweed is a sustainable source of nutrients for human consumption, and a rapid and accurate quality control approach is essential for ensuring the nutritional value and health benefits of seaweed. The nutritional constituents and quality of seaweed vary significantly along with their growth stages. The present study developed a quality control approach for seaweed Sargassum fusiforme, rapid and convenient, using near-infrared spectroscopy (NIRS) and chemometrics to predict protein, fat, and ash in S. fusiforme at different growth stages. Partial least squares (PLS) regression was utilized to construct the quantitative relationship between NIR spectral data and nutritional components. Moreover, three wavelength selection algorithms, namely, competitive adaptive reweighted sampling (CARS), genetic algorithm, backward interval (BI) and synergy interval, were implemented to optimize the PLS models and improve prediction accuracy. Results demonstrated that the CARS-PLS model exhibited superior performance with a root mean square error of prediction and a coefficient of determination of prediction of 0.1075 % and 0.9936 for protein, 0.1807 % and 0.7706 for fat, and 0.3315 % and 0.9937 for ash, respectively. The effects of different developmental stages on the nutritional quality of S. fusiforme were also investigated. The seedling and early growth stages were the preferable harvest times for preparing high-protein foods and health supplements. The overall results confirmed the strong applicability of NIRS as a rapid and accurate method for measuring the protein, fat, and ash contents of S. fusiforme. This work also recommended that S. fusiforme quality can be controlled in a more precise and refined manner on the basis of a customer-oriented strategy.