Urban　spatial　interaction　serves　as　an　indicative　measure　for　estimating　the　intensity　and　character　of　interurban　linkages　and　relationships.　The　previous　studies　have　utilized　intercity　relational　data　(e.g.,　population　migration,　goods　trade,　and　information　exchange)　to　build　up　urban　connections　directly.　Besides,　the　nighttime　light　(NTL)　data　have　also　been　adopted　to　simulate　a　dynamic　urban　intercity　flow.　However,　the　relevant　studies　have　not　clearly　defined　the　urban　spatial　interaction　based　on　the　NTL　data.　To　answer　this　question,　we　used　trial-and-error　testing　to　define　the　NTL-based　interaction.　First,　five　traditional　urban　interactions　were　selected　as　the　potential　definitions,　including　population　migration,　transfer　of　innovation,　information　flow,　financial　flow,　and　urban　composite　interaction.　Second,　as　usual,　the　NTL-based　urban　interaction,　named　the　Nighttime　Light　Interaction　(NTLI)　index,　was　simulated　based　on　the　NPP-VIIRS-like　NTL　data　and　the　radiation　model.　Taking　the　Yangtze　River　Delta　Urban　Agglomerations　(YRDUA)　as　an　example,　we　found　that　the　NTL-based　urban　interaction　is　more　like　the　population　migration　at　the　urban　agglomeration　scale　and　the　provincial　scale　with　R2　of　0.71　and　0.59,　respectively.　In　addition　to　this,　the　NTLI　index　has　a　weak　correlation　with　the　Transfer　of　Patent　(TP)　index,　Information　Flow　(IF)　index,　Economic　Interaction　(EI)　index,　and　Composite　Interaction　(CI)　index.　To　sum　up,　the　interaction　network　from　NTL　data　can　be　an　adequate　proxy　of　urban　population　interaction,　rather　than　the　knowledge　network　or　economic　network.　This　study　provides　a　new　thought　for　urban　network　simulation　and　urban　population　mobility　research.　IEEE