Infrared spectroscopy is a rapid, easy‐to‐operate, label‐free and therefore cost‐effective technique. Many studies performed on biofluids (eg, serum, plasma, urine, sputum, bile and cerebrospinal fluid) have demonstrated its promising application as a clinical diagnostic tool. Given all these characteristics, infrared spectroscopy appears to be an ideal candidate to be implemented into the clinics. However, before considering its translation, a clear effort is needed to standardise protocols for biofluid spectroscopic analysis. To reach this goal, careful investigations to identify and track errors that can occur during the pre‐analytical phase is a crucial step. Here, we report for the first time, results of investigations into pre‐analytical factors that can affect the quality of the spectral data acquired on serum and plasma, such as the impact of long‐term freezing time storage of samples as well as the month‐to‐month reproducibility of the spectroscopic analysis. The spectral data discrimination has revealed to be majorly impacted by a residual water content variation in serum and plasma dried samples.