Introduction:　To　promote　the　safety　level　of　ride-hailing　services,　this　study　develops　the　Targeted　and　Differentiated　Optimization　Method　of　Risky　Driving　Behavior　Education　and　Training　(TDOM-RDBET)　founded　on　driver　type　classification　of　high-risk　drivers.　Method:　Based　on　value　and　goal　orientations,　689　drivers　were　classified　into　four　driver　types　and　were　assigned　to　three　groups,　including　an　exper-imental　group,　a　blank　control　group,　and　a　general　control　group.　This　research　preliminarily　analyzes　the　effectiveness　of　the　TDOM-RDBET　to　reduce　mobile　phone　use　while　driving　by　assessing　the　main　effects　of　the　group　and　test　session　on　the　risk　value　ranking　of　mobile　phone　use　while　driving　(AR),　the　frequency　per　100　km　of　mobile　phone　use　while　driving　(AF),　and　the　frequency　per　100　km　of　risky　driving　behaviors　(AFR),　as　well　as　the　interactive　effects　of　the　two　factors　on　AR,　AF,　and　AFR,　based　on　a　two-way　analysis　of　variance　(two-way　ANOVA).　Results:　The　results　demonstrate　an　overall　significant　reduction　in　AR　(F　=　8.653,　p　=　0.003),　AF　(F　=11.027,　p　=　0.001),　and　AFR　(F　=　8.072,　p　=　0.005)　for　the　experimental　group　after　training.　Moreover,　significant　interactive　effects　of　the　driver　group　x　test　ses-sion　on　AR　(F　=　7.481,　p　=　0.001)　and　AF　(F　=15.217,　p　＜　0.001)　were　found.　AR　was　significantly　lower　for　the　experimental　group　than　for　the　blank　control　group　(p　＜　0.05)　in　the　post-training　condition.　Moreover,　AF　was　also　significantly　lower　for　the　experimental　group　than　for　the　blank　control　group　(p　＜　0.05)　and　general　control　group　(p　＜　0.05)　in　the　post-training　condition.　Practical　Applications:　On　the　whole,　it　was　preliminarily　verified　that　the　TDOM-RDBET　is　more　effective　than　the　general　training　method　at　modifying　the　risky　driving　behavior.　&　COPY;　2023　National　Safety　Council　and　Elsevier　Ltd.　All　rights　reserved.