ABSTRACT.- Long-term agricultural experiments (LTE) are essential to detect cumulative treatment effects on soil properties and to design sustainable production systems. However, to ensure high quality of long-term soil data and their correct interpretation, several analytical issues regarding the accuracy and analytical laboratory bias need to be considered. This paper aims to (1) evaluate laboratory precision and trueness of analytical soil data for the evaluation of long-term trends in LTE and to (2) assess the limitations and challenges for non-certified soil laboratories that might compromise the quality of analytical soil data. A data set of internal reference soil materials (IRM) collected over 16?years and interlaboratory data from eleven years were analyzed to verify method precision, trueness, and the subsequent long-term dataset reliability for several soil quality parameters: organic carbon (SOC, determined either by wet or dry combustion), pH (water), extractable phosphorous (either Bray I or citric acid; Bray-P or citric acid-P), and exchangeable potassium (Kexch). Results showed that IRM used by the laboratory were homogenous in terms of physical and chemical composition and appropriate to confirm the precision of long-term soil survey data. The relative standard deviation for repeatability and reproducibility (RSDR) ranged from 1.5% for SOC by wet combustion to 9.5% for citric acid-P. HorRat values (the ratio of the estimated standard deviations of reproducibility and the repeatability found for individual analytical procedures) for all chemical soil properties were within the acceptable ranges of