Two different methods are used to determine the limits of uncertainty for environmental measurements. The first employs the general propagation of error formula. In this method, the precisions and accuracies of each of the variables of the mathematical expression defining the final measurement result are used. A second method utilizes the agreement of results from co-located samplers to estimate precision and results of independent audits to estimate bias. Each of the methods has certain advantages. The end results, using the dichotomous particulate sampler as an example, show that the use of co-located samplers reveals the presence of unaccounted for sources of error not considered by the method using the propagation of error formula.