Abstract: A Bayesian analysis method is given when two parameters μ and σ² are unknown at the present of the lognormal distribution. We give two typies of jount conjugale prior distribution for μ and σ². When the life of product abides by the lognormal distribution,for the first type of the first type of the joint conjugate prior,we give the necessary and sufficient condition for the existence of the posterior mean and variance of the life of product, then we give the estimator for the mean life of product by the posterior mean for two typies of joint conjugate prior distribution.two corresponding examples are given in this paper.Meantime we get the Asymptotic posterior normal distribution for large sample,which extends the corresponding result in 2