In order to undertake the simulation and prediction of ENSO, based on the global oceanic general circulation model with lower resolution developed by the Institute of Atmospheric Physics, Chinese Academy of Sciences, after the introductions of Richardson number dependent vertical diffusion and the physical process of solar short-wave radiation penetration, an oceanic general circulation model of upper tropical Pacific with higher resolution was developed. The numerical simulation studies on the structure and evolution of the temperature and circulation in the tropical Pacific Ocean were performed with this model and the observed atmospheric forcing fields from 1980 to 1995. And focusing on the three dimensional characteristics of ocean and its evolution, the simulated results were validated by using the oceanic assimilated analysis during the same period produced by National Center of Environmental Prediction (NCEP) in USA. At first, the ability of this model to simulate the three dimensional features of the structure of ENSO and their evolution was investigated. It was indicated that: basically, the creation, development and decay of all cold and warm events during the 16 years were modeled correctly. There were some errors in the strength and structure of oceanic temperature anomalies, especially, at the subsurface, in the western equatorial Pacific and along the thermocline, the temperature anomalies were significantly weaker than the observed ones. At the sea surface, the oceanic temperature (SST) anomalies were closer to observation and only larger than the observed in the vicinity of dateline. Then, it was emphasized that, the correct simulation of SST was not enough for the successes of atmosphere-ocean coupled model and the prediction of ENSO, the really stringent was the capabilities of oceanic model simulating the ocean at the subsurface. Therefore, in order to evaluate the oceanic model more comprehensively and approach the reasons of model’s errors, after the regions with the most significant interannual and seasonal variability were found out based on the NCEP oceanic analysis, some statistical amounts, i. e., multiyear averaged state and its seasonal variations, interannual variability and its seasonal variations, were validated. The results indicated that the main features of statistical annual mean states of oceanic temperature and current were simulated correctly, but the errors were also obvious, particularly, there were large systematic errors in the simulations of some oceanic circulation systems at subsurface such as thermocline, equatorial trough, northern equatorial ridge, northern equatorial counter-current trough and equatorial under-current and so on. It was these errors that determine the spatial distributions of the modeling errors of other statistical amounts.