Nonlinear　microwave　device　modeling　is　an　important　part　of　computer-Aided　design　(CAD)　and　many　papers　have　been　published　in　the　literature.　This　paper　presents　a　review　of　recent　neural　network　approaches　to　the　modeling　of　nonlinear　microwave　device　including　the　dynamic　Neuro-SM　approach　and　the　Wiener-Type　dynamic　neural　network　approach　and　its　applications.　DC,　small-signal　and　large-signal　harmonic　data　are　used　as　training　data.　The　neural　network　based　methods　can　fast　and　accurately　build　accurate　models　for　nonlinear　microwave　devices.　Compared　with　conventional　equivalent　circuit　models,　the　models　generated　by　these　neural　network　based　methods　are　more　efficient　to　represent　the　behavior　of　the　device.　©　2020　IEEE.