A　computer　software　package,　SAVFS,　is　presented　for　statistical　analysis　of　vertical　facies　sequences　using　the　Markov　chain　or　variations　of　it.　The　analysis　may　reveal　the　possible　presence　of　preferred　vertical　occurrences　of　facies　in　sedimentary　successions.　SAVFS　is　written　in　FORTRAN　77　for　IBM-PC　or　compatible　microcomputers.　Based　on　the　combination　of　some　generally　used　methods,　SAVFS　provides　three　analytical　methods,　Gingerich-Harper,　Powers-Easterling　and　Selley-　Turk.　This　remedies　some　of　the　problems　occurring　with　single　methods　and　improves　the　statistical　techniques.　SAVFS　calculates　various　matrices　and　automatically　estimates　significance　levels.　In　the　Gingerich-Harper　method,　SAVFS　allows　the　user　to　test　significance　at　different　levels.　In　addition,　the　package　performs　a　comparison　of　multi-sections　so　that　data　from　several　similar　stratigraphic　sections　may　be　combined　and　presented　as　a　single　section.　A　set　of　scales　for　significance　derived　from　the　values　of　the　normalized　difference　and　binomial　probability　matrices　is　presented.　This　makes　the　results　of　the　Powers-Easterling　and　Selley-Turk　methods　comparable　with　those　of　the　Gingerich-Harper　method.　A　case　study　of　the　Courtmacsherry　Formation　(Tournaisian)　in　Ireland　using　SAVFS　demonstrates　that　two　facies　transition　successions,　sandy　and　muddy　successions,　are　revealed.　The　sandy　facies　transition　succession　well　illustrates　a　coarsening-upward　sequence,　representing　a　storm-and　wave-dominated　shallow　marine　environment.　The　vertical　facies　sequences　of　the　formation　are　extracted　from　this　apparently　random　succession.　(C)　1998　Elsevier　Science　Ltd.　All　rights　reserved.