A humus sample and a mineral soil sample were collected to investigate different sample pretreatments and their effects on the quality (mean value and uncertainty) of soil chemical data. Thirty-four humus subsamples and sixty mineral subsamples were prepared for chemical analysis. Diverse combinations of sieving, milling or grinding were utilized before the determination of Kjeldahl nitrogen (Kj-N), pH, extractable elements, and a few other chemical parameters. The humus subsamples were sieved, sieved and milled or sieved and ground. Relative to sieving, grinding improved the uncertainty in the determination by 2–10 times [highest for calcium (Ca), manganese (Mn), Kj-N, loss on ignition and volume weight]. The mineral soil subsamples were either sieved or sieved and ground. The uncertainty was improved by 2–4 times for aluminum (Al), barium (Ba), Ca, hydrogen (H), potassium (K), sulphur (S), silicon (Si), zinc (Zn), and nitrogen (N) when grinding was applied. The result increased for almost all of the chemical parameters when the samples (humus and mineral) were ground in addition to sieving. The maximum increase in mean value was observed for Si (330%), iron (Fe) (200%) and carbon (C), K, magnesium (Mg), Mn, sodium (Na) (17–34%). The determination of Kj-N was less influenced (0–16%) by different pretreatments but the uncertainty in the determination was improved up to 10 times when the sample was ground. Different sieving techniques were also utilized. It was shown that different sieving techniques altered the mean value for Ba, Fe, K, Mg, Si, Zn, pH, and loss on ignition, while the uncertainty was unaffected. All subsamples were reanalyzed after 15 months, without mixing in each sample before weighing. An increase in concentration was observed for C, Fe, phosphorus (P), S (up to 130%) and in volume weight (6–26%) while a decrease was observed for Ca, K, Mg, strontium (Sr), Zn (7–27%).