Quantification　of　cracking　is　crucial　to　assessing　overall　cracking　condition,　monitoring　cracking　propagation,　and　other　pavement　design　and　managerial　decision-making　purposes.　For　a　majority　of　the　highway　agencies,　the　cracking　width　and　length　are　the　two　most　important　indicators　representing　severity　and　extent　of　cracking,　respectively.　However,　due　to　limitations　in　pavement　data　collection　technology　and　vague　protocols,　cracking　length　and　width　were　mostly　estimated　or　simply　summed　or　averaged　instead　of　accurately　measured　over　the　past　two　decades.　In　recent　years,　accurate　cracking　map　can　be　derived　in　a　consistent　manner　with　the　advance　of　the　three-dimensional　(3D)　data　collection　systems.　Nonetheless,　rare　effort　has　been　dedicated　to　rigorous　cracking　length　and　width　measurement,　which　has　slowed　the　progress　of　cracking　measurement.　Although　the　AASHTO　PP67-10　protocol　specifies　new　requirements　for　automated　cracking　measurement,　cracking　quantification　remains　immature　in　current　practice,　which　affects　the　reliability　of　pavement　maintenance　and　design　modeling.　In　this　research,　a　rigorous　methodological　framework　for　accurate　cracking　length　and　width　measurement　is　developed.　Methods　are　introduced　and　demonstrated　effective　at　extracting　the　desirable　cracking　skeleton,　which　is　used　for　accurate　cracking　length　measurement.　Based　on　the　cracking　skeleton　and　contour,　an　orthogonal　projection　method　is　developed　in　this　study　to　calculate　the　continuous　cracking　width　for　each　pixel　along　the　cracking　skeleton.　This　study　provides　insight　into　how　to　derive　reliable　and　consistent　cracking　data　from　3D　images.　Findings　from　this　research　would　further　improve　the　practice　of　cracking　measurement　based　on　the　new　AASHTO　PP67-10　protocol.　(C)　2016　American　Society　of　Civil　Engineers.