Fumonisins are mycotoxins produced by Fusarium species. Their contamination of grain and animal feed poses significant health risks to both humans and livestock. Among the fumonisins, FB1, FB2, and FB3 are commonly found in staple crops like maize, rice, and wheat, necessitating reliable detection and quantification methods to ensure food and feed safety. We successfully developed and validated a rapid and optimised method for the simultaneous detection and quantification of three major fumonisins in diverse grain and feed matrices (rice, wheat flour, maize, alfalfa, oat grass, and soybean meal) by LC-MS/MS. Various extraction techniques and clean-up strategies have been evaluated to address the challenges posed by complex sample matrices. Finally, an acetonitrile/water/formic acid solution (50:49:1, v/v/v) and the multi-plug filtration clean-up (m-PFC) column were selected for fumonisins extraction and clean-up respectively, due to their ability to offer superior performance in terms of both recovery rates and minimising interference. The results of matrix effects demonstrated that solvent-based calibration curves were adequate for simpler matrices such as rice and wheat flour, while matrix-matched calibration was essential for more complex matrices like alfalfa and oat grass. Recovery rates for fumonisins in various matrices ranged from 72.5% to 99.3%, while the limits of detection and quantification ranged from 0.3 to 3.3 µg/kg and 1 to 10 µg/kg, respectively, indicating the robustness and reliability of the method. This optimised approach provides a cost-effective and highly reliable solution for fumonisins monitoring, thereby helping mitigate the health risks associated with mycotoxins in food and feed products.