Abstract: Since the existing traffic conflict techniques are not applicable to identify serious conflicts between vehi-cles and non-motor vehicles in China.And drivers'passing behaviors through the intersections are proven to play a signifi-cant role in the occurrence of traffic collisions.The study proposed a method of identifying serious conflicts between vehi-cles and non-motor vehicles in consideration of drivers'passing behaviors.The traffic scene videos were pre-processed by applying coordinate transformation technique,then the conflict data of vehicles and non-motor vehicles at the intersections was extracted.Critical time difference to collision (CTDTC)was proposed to depict drivers'passing behaviors.The results show that the difference of drivers'passing behaviors increases,which tend to trigger collisions,when the absolute value of TDTC is less than 0.66 s.The test results also indicate that the cumulative frequency of non-serious conflicts is about 1 5% and 85% for serious conflicts,which are confirmed by a cumulative frequency method,when 0.66 s of TDTC is de-fined as an identification threshold.The proposed method defines a specific threshold of identifying serious conflicts be-tween vehicles and non-motor vehicles and provides more accurate and objective results.