This paper selectively reviews various approaches of macroeconometric modelling and highlights some important lessons from more than half a century of model-building particularly in the context of Asian countries. Addressing several issues discussed in this paper can improve the use of macroeconometric models (MEM) in forecasting and policy analysis in the foreseeable future. This survey shows that most MEMs in developing countries are either becoming smaller in size or not being subject to a thorough diagnostic investigation. In the specification of models one should consider the interplay among macroeconomic policies of different countries via international trade and global financial markets. It is argued that the Project Link and the Fair multi-country model are two initiatives in the right direction. It also appears that with advancement of econometric "know-how", the disparity of opinions between advocates and critics of macroeconometric modelling can be narrowed.