Three long standing problems in electron transport in a mesoscopic system are explanations for the nonlinear conductances observed in a quantum point contact, various systems of scanning tunneling microscopy (STM), and a single-electron transistor with a quantum dot. Peak splitting by magnetic field is the issue in the last system. These Kondo-related transport phenomena are classified as the steady-state Kondo phenomena. In this talk, I briefly review experimental results for these systems and discuss about new transport mechanism through the resonant tunneling levels (RTLs), which are formed only when a bias is applied. With this new transport mechanism and the transport by Kondo effect, we explain the lineshapes of nonlinear conductance for the entire range of applied bias used in the experiments. Interestingly, in the STM systems that show asymmetric lineshapes, Kondo effect is negligible and the lineshape is mostly determined by the RTL peaks.