Metabolomics has been successfully applied to guide the rational engineering of industrial strains and improve the performance of bioprocesses. Mortierella alpina has traditionally been one of the most popular industrial strains for the production of polyunsaturated fatty acids. However, a systematic comparison and optimisation of the metabolomic analysis methods of M. alpina has not yet been reported. We sought to identify potential weaknesses that are important for accurate metabolomic analysis. We also aimed to determine an efficient sample preparation protocol for metabolomics studies in the oleaginous filamentous fungus M. alpina. In this study, using GC-MS, we evaluated three sample preparation protocols and five solvent mixtures by assessment of the metabolite profile differences, the sum of peak intensities and the reproducibility of metabolite quantification. The freeze-dried biomass had better reproducibility and recovery than fresh biomass for metabolite extraction and data normalisation that is part of a metabolomics analysis of filamentous fungi M. alpina. Methanol:water (1:1) was superior for the profiling of metabolites in oleaginous fungi M. alpina. The unbiased metabolite profiling difference between the growth phase and lipids synthesis phase revealed that the degradation of amino acids were critical nodes for the efficient synthesis of lipids in M. alpina. The use of freeze-dried biomass for metabolite extraction and data normalisation was more efficient at measuring the active state of the intracellular metabolites in M. alpina. We recommend extracting the intracellular metabolites with methanol:water (1:1). An important role of amino acid oxidation in the nitrogen limitation-mediated lipid accumulation was found.