Abstract: Virtual population analysis (VPA) is one of the classic models in fish stock assessment, and is one of the main models in calculating the total allowable catch (TAC). Because of the indeterminacy of VPA, the catch-at-age data itself can not estimate the fishing mortality (and stock size) of the terminal years, therefore it requires the catch-at-age data and the auxiliary data. The tuning methods in VPA originated from Europe, they are simple to understand and do not require complicated calculations, and are widely applied in the practical fish stock assessment. This paper analyzed the three most typical tuning methods: JAM(judicious average method), LS(Laurec and Shepherd method), Hybrid. Simulation analysis shows that when fishing mortality rates are in status quo, all the three methods can produce correct solutions. When fishing mortality rates change with time, the LS method has better performances than the other two.