Hazardous　materials　are　harmful　to　environment　and　human　health　due　to　their　toxic　ingredients.　It　is　very　important　to　intensify　efforts　in　the　safety　management　of　hazardous　materials,　and　fundamentally　prevent　and　reduce　accidents.　In　this　study,　we　consider　a　time-dependent　hazardous　materials　vehicle　routing　problem　in　a　two-echelon　supply　chain　system.　The　goal　is　to　determine　the　departure　time　and　the　optimal　route　with　a　minimum　risk　value　for　hazardous　materials　transportation.　Considering　that　time　has　a　signicant　influence　on　the　transportation　risk,　we　formulate　a　time-dependent　transportation　risk　model　and　propose　a　credibilistic　mixed　integer　programming　model　to　minimize　the　expected　risk.　An　improved　genetic　algorithm　whose　chromosomes　contain　two　types　of　genes　is　designed　to　handle　the　proposed　model.　Numerical　experiment　is　given　to　illustrate　the　efficiency　of　the　proposed　model　and　algorithm.　Compared　with　the　traditional　transportation　risk　model,　the　time-dependent　transportation　risk　model　can　signicantly　reduce　the　risk　around　42%.　©　2017　World　Academic　Press,　UK.　All　rights　reserved.