This study reports total nitrogen (TN), carbon (TC) and organic carbon (TOC) contents, as well as stable isotopic ratios (δ15NTN, δ13CTC and δ13Corg), from marine, stream and lacustrine geological reference materials (JMs‐1, JMs‐2, SBC‐1, SCo‐2, JSd‐3, JSd‐4 and JLk‐1), which can be classified into two types; modern surface, and ancient, sediments. Using this newly obtained data set, we systematically investigated the biases of three pre‐treatment methods for nitrogen and carbon measurements: (1) non‐acidification (untreated method), (2) HCl acidification in various conditions and subsequent rinsing with high‐purity water repeatedly (rinse method) and (3) in situ acidification in a gold‐coated silver capsule (capsule method). Our results showed that such acid pre‐treatments can affect both nitrogen and carbon measurements, except for the carbon data of the ancient geological RMs. Therefore, the nitrogen analysis should be conducted using untreated samples while carbon data should be obtained using the capsule method for modern surface sediments and the rinse method (6 mol l‐1 HCl acidification at 70 °C) for ancient sediments. We believe that using appropriate pre‐treatment methods is crucial to deriving new interpretations of ancient/modern environmental processes and mechanisms in the future.