In　order　to　evaluate　and　study　the　influence　of　smog　environment　on　driving　safety,　this　paper　utilizes　the　measurability　of　mood　states,　adopts　back　propagation　(BP)　artificial　neural　network　instrument　to　establish　a　smog-risky　mood　network　model.　Six　pictures　in　different　smog　levels　were　used　as　emotional　stimuli　to　verify　the　BPM　neural　network-based　smog-risky　mood　relation　model.　The　result　shows　that　the　smog　is　a　main　factor　affecting　the　driver＇s　mood.　The　method　of　estimating　moods　based　on　the　smog　environment　of　BP　network　is　able　to　predict　70%　of　the　danger.　It　was　proved　to　be　an　effective　means　for　the　safety　management　and　self-detection　of　drivers,　especially　professional　drivers.