Energy　efficiency　of　train　operations　is　influenced　largely　by　the　speed　control　and　the　scheduled　running　time　in　the　train　timetable.　In　practice,　the　running　time　of　a　train　is　often　determined　in　the　train　timetabling　process　at　the　macroscopic　level　while　the　energy-efficient　speed　control　of　a　train　on　a　segment　is　often　determined　at　the　microscopic　level　with　the　given　timetable.　They　are　usually　optimized　separately　due　to　limited　computational　resources,　which　however　may　result　in　sub-optimal　solutions.　To　address　this　issue,　this　paper　proposes　a　novel　integrated　micro-macro　approach　for　better　incorporating　train　energy-efficient　speed　control　into　the　railway　timetabling　process.　Firstly,　we　formulated　the　integrated　train　timetabling　and　speed　control　optimization　problem　as　a　nonlinear　mixed-integer　programming　model.　Due　to　its　complexity,　we　reformulate　it　on　the　basis　of　flow　conservation　theory　in　a　space-time-speed　(STS)　network　and　solve　the　problem　in　two　steps.　In　the　first　step,　a　set　of　pre-solved　energy-efficient　train　trajectory　templates　is　generated　by　a　segment-level　optimization　approach　with　consideration　of　train　travel　time,　entry　speed　and　exit　speed　to　save　computation　time.　In　the　second　step,　a　near-optimum　train　energy-efficient　timetable　solution　is　found　by　a　fast　algorithm,　which　consists　of　the　shortest　generalized　cost　path　algorithm,　conflict　detection　and　resolution　algorithm,　and　calculation　of　dynamic　headways　between　two　successive　trains.　The　numerical　experiments　demonstrate　that　the　developed　approach　provides　better　outcomes　than　the　benchmark　case　in　terms　of　both　train　journey　time　and　energy　consumption.