This　paper　proposes　a　correlation　coefficient　maximization　approach　(CCMA)　for　estimating　priorities　from　a　pairwise　comparison　matrix.　The　priorities　are　supposed　to　be　as　highly　correlated　with　each　column　of　a　pairwise　comparison　matrix　as　possible.　Such　priorities　are　not　unique　and　can　be　determined　in　different　ways.　Two　optimization　models　are　therefore　suggested　for　determining　further　the　priorities,　one　of　which　leads　to　an　analytic　solution.　Theorems　about　the　CCMA　are　developed　and　its　potential　applications　are　illustrated　with　two　numerical　examples.　©　2006　Elsevier　Inc.　All　rights　reserved.