AbstractThe univariate fatigue life distribution proposed by Birnbaum and Saunders [A new family of life distributions. J Appl Probab. 1969;6:319–327] has been used quite effectively to model times to failure for materials subject to fatigue and for modelling lifetime data and reliability problems. In this article, we introduce a Birnbaum–Saunders (BS) distribution in the multivariate setting. The new multivariate model arises in the context of conditionally specified distributions. The proposed multivariate model is an absolutely continuous distribution whose marginals are univariate BS distributions. General properties of the multivariate BS distribution are derived and the estimation of the unknown parameters by maximum likelihood is discussed. Further, the Fisher's information matrix is determined. Applications to real data of the proposed multivariate distribution are provided for illustrative purposes.Keywords: Birnbaum–Saunders distributionconditional specificationsfatigue life distributionlifetime datamultivariate distributions AcknowledgementsArtur Lemonte gratefully acknowledges grants from CNPq (Brazil) and FACEPE (Pernambuco, Brazil). Germán Moreno–Arenas gratefully acknowledges grants from Mobility Program of the Universidad Industrial de Santander (Colombia). The authors would like to thank two anonymous referees for valuable comments.