With　the　rise　of　the　transportation　mode　of　shared　electric　bikes　(shared　e-bikes)　in　China,　shared　e-bike　related　accidents　have　gradually　increased.　To　facilitate　the　design　of　safety　policies,　it　is　important　to　understand　the　factors　that　influence　shared　e-bike　riders＇　traffic　accidents　to　facilitate　intervention　strategies.　For　this　purpose,　the　structural　equation　model　(SEM)　with　mediation　analysis　was　applied　by　incorporating　seven　latent　factors:　traffic　accidents,　traffic　violation　behaviors,　attitude　toward　safety　responsibility,　and　attitude　toward　rule　violations,　risk　perception,　perceptive-motor　skills,　and　safety　skills.　A　questionnaire　survey　of　a　sample　of　406　shared　e-bike　riders　in　China　was　conducted　to　obtain　self-reported　survey　data.　The　results　reveal　that　traffic　violation　behaviors　and　attitude　toward　safety　responsibility　had　a　statistically　significant　consequence　on　traffic　accidents.　Attitude　toward　rule　violations,　perceptive-motor　skills,　and　safety　skills　can　predict　shared　e-bike　riders＇　traffic　accidents　when　the　traffic　violation　behaviors　are　used　as　a　mediator.　Moreover,　risk　perception　could　also　be　used　to　predict　shared　e-bike　riders＇　traffic　accidents　when　using　attitudes　toward　safety　responsibility　or　rule　violations　and　traffic　violation　behaviors　as　a　mediator.　This　paper　lays　a　foundation　for　policymakers　and　traffic　managers　to　develop　effective　intervention　strategies　and　improve　shared　e-bike　safety.