Four aspects of uncertainty linked to the laboratory of medical analysis are discussed. 1) Uncertainty and the laboratory test: the post-test probability of a diagnostic is intimately linked to the ability to establish clinically the pre-test probability. 2) Uncertainty and the patient: reference values are most often chosen from the literature. The analytical methodology as well as the population of reference should be carefully checked. 3) Uncertainty and the laboratory result: this uncertainty will diminish if preanalytical conditions are standardized and if analytical imprecision is known. The analytical imprecision is given by the coefficient of variation (CV) of the internal quality control. The CV is used to calculate the critical difference. 4) Uncertainty and the practitioner: to efficiently diminish this uncertainty in the case of unusual questions of the medical practice, a network should be established between the practitioner, the medical specialists and the scientific specialists of the laboratory.