Traditional　traffic　lights　control　system　works　on　a　fixed-time　basis,　which　can＇t　optimize　the　time　of　traffic　lights　according　to　the　change　of　traffic　flow.　To　address　this　issue,　this　paper　proposes　an　approach　based　on　data-driven　belief　rule　base　for　smart　traffic　lights.　The　main　idea　of　this　approach　is　using　historical　traffic　data　to　predict　traffic　flow,　and　then　determine　traffic　lights　control　strategy　in　term　of　predicted　traffic　flow　and　road　conditions.　Utilizing　the　data-driven　belief　rule　base　to　forecast　traffic　flow　is　proposed　firstly.　It　is　worth　nothing　that　the　data-driven　belief　rule　base　can　transform　data　into　rules　directly　involving　neither　time-consuming　iterative　learning　procedure　nor　complicated　rule　generation　mechanisms.　Subsequently,　a　simple　but　efficient　method　for　controlling　traffic　lights　based　on　priority　is　proposed.　Then　some　case　studies　are　provided　to　illustrate　the　feasibility　and　efficiency　of　the　proposed　approach.