The　objective　of　this　paper　is　to　test　the　effectiveness　and　adaptability　of　speed　reduction　markings　(SRMs)　in　downhill　sections　on　urban　roads　with　distinct　roadway　grades.　Empirical　data　including　vehicle　speed　and　acceleration　were　collected　in　a　driving　simulator.　Subjective　questionnaires　were　conducted,　and　two　indexes　-　the　relative　speed　difference　and　standard　deviation　of　acceleration　-　were　developed　to　evaluate　the　effectiveness　and　adaptability　of　SRMs.　Meanwhile,　the　effectiveness　of　driving　simulator　related　to　different　road　alignments　and　types　of　SRMs　has　been　validated　through　a　field　test.　Results　of　subjective　questionnaires　showed　that　the　majority　of　subjects　had　no　feelings　of　nervousness,　but　they　were　affected　by　SRMs　while　driving　through　downhill　sections　in　all　four　scenarios　(i.e.,　downhill　sections　with　vertical　grades　of　3,　2,　1.5　and　1%).　In　terms　of　vehicle　speed　and　acceleration,　the　results　of　the　analysis　of　variance　(ANOVA)　and　the　contrast　analysis　(S-N-K　method)　indicated　that　SRMs　were　significantly　effective　when　roadway　grades　of　downgrade　sections　were　1.5,　2　and　3%,　while　transverse　speed　reduction　markings　(TSRMs)　had　significantly　worse　adaptability　(P　＜　0.05).　Therefore,　this　research　recommends　that　TSRMs　could　be　placed　in　downhill　sections　with　roadway　grades　of　1.5　or　2%;　longitudinal　speed　reduction　markings　(LSRMs)　could　be　placed　in　downhill　sections　with　a　roadway　grade　of　3%.　Whether　SRMs　are　placed　in　downhill　sections　with　a　roadway　grade　of　1%　would　depend　on　other　factors　such　as　financial　issues　and　crash　records,　which　are　not　considered　in　this　paper.　(C)　2014　Elsevier　Ltd.　All　rights　reserved.