Abstract: In this paper,some special problems of objective analysis for the tropics,such as sparsity of observational data,systematic error of the first guesses,balance between wind and mass fields,and analyses of divergent component of wind,moisture and surface state are discussed. Some possible ways to cope with these problems are also investigated.