This paper is devoted to the study of the influence of experimental parameters and calculation hypotheses on the estimated value of Weibull's modulus, m, of structural ceramics. Numerical simulation programs have been written and rupture tests have been performed in order to characterize the role of the number of samples tested and thus of the probability estimator. One can thus define an optimal value of the number of samples needed to estimate Weibull's modulus with a given uncertainty. Other numerical programs simulate the effects of the loading rate as well as the effects of Paris' law constant, A and the propagation exponent, n, on the m value.