The　group-ranking　problem　is　ubiquitous　in　the　real　life,　which　is　a　group　decision-making　(GDM)　problem　and　cannot　be　completed　independently　by　individuals.　However,　different　experts　or　decision-makers　(DMs)　may　provide　their　preference　information　on　alternatives　in　the　form　of　uncertain　preference　ordinals.　This　paper　developed　an　evidential　reasoning　(ER)　based　method　to　deal　with　various　types　of　preference　information　on　alternatives　such　as　precise　and　imprecise,　complete　and　incomplete,　and　known　and　unknown,　which　may　provide　by　experts　or　DMs　in　the　process　of　alternatives　ranking.　The　proposed　method　allows　experts　or　DMs　to　express　their　preferences　independently　and　freely　using　belief　structure,　and　provides　a　means　based　on　evidence　distance　to　determine　the　relative　weights　of　belief　structures,　which　is　not　need　to　solve　the　complex　optimization　model.　Furthermore,　the　interval　ER　algorithm　is　employed　to　aggregate　different　types　of　preference　information　with　a　rigorous　and　systematic　framework.　The　feasibility　and　rationality　of　the　method　are　explained　and　verified　by　examples.　©　2019　-　IOS　Press　and　the　authors.　All　rights　reserved.