The　random　combination　of　cation　and　anion　makes　ionic　liquids　to　be　“designer　solvents,　＂　meanwhile　at　least　a　million　binary　ILs　can　potentially　be　constructed　by　adjusting　the　anion,　length,　position,　or　the　number　of　alkyl　chain　on　the　organic　cation.　It　is　tedious,　time-consuming,　and　almost　impossible　to　measure　the　properties　of　all　ILs　by　experiment.　Thus,　predictive　thermodynamic　models　are　urgently　needed　for　the　tailor-making　of　ionic　liquids　for　specific　tasks.　In　this　chapter,　the　state　of　the　art　of　famous　and　widely　used　predictive　thermodynamic　approaches　for　ionic　liquids　is　classified　into　two　types:　activity　coefficient　models　(COSMO-based　model,　UNIFAC-based　model,　QSPR/QSAR　model,　and　RST　model)　and　equation　of　state　(EOS)　models　(SAFT-based　EOS,　GC　EOS,　and　lattice　models),　are　addressed.　©　2021　Elsevier　Inc.　All　rights　reserved.