A Class of Bivariate Models with Proportional Reversed Hazard Marginals Debasis Kundu and Rameshwar D. Gupta Abstract Recently the proportional reversed hazard model has received a considerable amount of attention in the statistical literature. The main aim of this paper is to introduce a bivariate proportional reversed hazard model and discuss its different properties. In most of the cases the joint probability distribution function can be expressed in com- pact forms. The maximum likelihood estimators cannot be expressed in explicit forms in most of the cases. EM algorithm has been proposed to compute the maximum like- lihood estimators of the unknown parameters. For illustrative purposes two data sets have been analyzed and the performances are quite satisfactory. Keywords: Joint probability density function; Conditional probability density function; Maximum likelihood estimators; EM algorithm. Department of Mathematics and Statistics, Indian Institute of Technology Kanpur, Kanpur, Pin 208016, INDIA. Phone no. 91-512-2597141, Fax No. 91-512-2597500, e-mail: kundu@iitk.ac.in. Corresponding author. Part of his work has been supported by a grant from the Department of Science and Technology, Government of India. Department of Computer Science and Statistics, The University of New Brunswick at Saint John, New Brunswick, Canada E2L 4L5. Part of his work has been supported by a discovery grant from NSERC, CANADA. 1