Life time data are mainly used in medicine, public health, and engineering sciences. The classical statistics related to these fields are based on precise measurements. However, according to modern measurement science precise measurement of continuous phenomena is not possible. In addition the observed quantity is more or less precise. In order to obtain realistic results, in addition to classical and Bayesian inference, the most up to date model for this are fuzzy number approaches which are more suitable and realistic. In this study, I consider the classical and Bayesian inference criteria for the Pareto distribution to obtain more realistic results for fuzzy observations of life time. In addition to stochastic variation the proposed generalized estimators cover fuzziness of the life times as well.