The measurement process consists of a set of sub-processes where the beginning is in the selection of the sample until the end of the analytical steps. In a simplified way, it can be established that the uncertainty coming from the measurement process is constituted by the association of the uncertainty coming from the set of these sub-processes. The literature has organized this set of subprocesses into two important areas, the sampling process and the analytical process, each one with its own uncertainties characteristics. In order to better estimate the measurement uncertainty, it is necessary to fully consider the uncertainty arising from both processes without firstly determining which component will be most significant for the measurement uncertainty. This work project was motivated by the indifference to the sampling process, which is usually carried out by little experienced professionals, without knowledge of how much uncertainty of sampling can affect the measurement uncertainty. The objective of this dissertation was to evaluate how much uncertainty is the uncertainty of the measurement uncertainty in addition to stimulating the dissemination of the culture of this practice in physicochemical measurements; since, actually, only analytical uncertainty is privileged. The methodologies used were the analysis of the state of art of sampling uncertainty and the treatment of a data set by four different statistical techniques: classical variance analysis, robust variance analysis, and two different models of the Range Statistics. The results showed that there is no behavior pattern regarding to sampling uncertainty versus analytical uncertainty, as well as the absence of methodologies that take into consideration the characteristics of the data. It is concluded that the sampling uncertainty can impact the uncertainty of measurement, regardless of the matrix and / or quantity in question.