In　order　to　determine　the　random　factors　of　traffic　capacity,　so　as　to　calculate　the　signalized　intersection　capacity　accurately.　The　random　factors　were　divided　into　regional-specific　factors　and　intersection-specific　factors.　A　discrimination　method　for　modeling　the　random　factors　that　influence　signalized　intersection　capacity　was　proposed　based　on　the　average　headway.　The　random　factors　were　selected　using　the　extension　method　based　on　their　degrees　of　influence　and　the　probabilities　of　occurrence.　The　influence　level　of　random　factors　were　divided　to　low,　medium　and　high　according　to　the　degrees　of　influence　and　the　probabilities　of　occurrence.　Actual　data　from　25　intersections　located　in　two　cities　located　in　the　northern　and　southern　regions　of　China　were　used　to　verify　the　discrimination　method.　The　results　show　that　the　random　factors　can　reflect　typical　characteristics　of　the　city　or　an　intersection,　which　greatly　influence　signalized　intersection　capacity,　and　they　remain　stable　during　a　certain　period.　It　is　recommended　that　a　random　factor　should　be　considered　in　the　capacity　calculation　model　at　signalized　intersections　when　its　influence　level　is　medium　or　high.　©　2016,　Editorial　Department　of　China　Journal　of　Highway　and　Transport.　All　right　reserved.