In this article, we are interested in asymptotic properties of the maximum pseudolikelihood estimators of the parameters of a GPD model with interval censoring. A first step consists in an exact calculation of the probability distribution of the random vector (Y , Z, Δ) associated with the censored model by interval, and its probability density is defined. Then, the efficiency and the asymptotic normality of the maximum pseudolikelihood estimators are stated in two propositions. A second step, based on simulations and the Barzilai-Borwein algorithm, provides the efficiency and asymptotic normality of the maximum pseudo-likelihood estimators of the parameters of a GPD model.