In view of the 60th anniversary of the pioneering oceanographic work on the numerical modeling of dynamical characteristics of the ocean [1], the author has decided to present his view of the principal mile stones in the progress of the numerical modeling of climatic characteristics of the ocean. This progress is shown schematically and conventionally as a single a table consisting of two lines: (A) synthesis of measure ment data models and (B) theory and calculation of sea currents. Line A consists of the following stages: ref erence surface method, diagnostic method, diagnostics-adaptation, and four dimensional analysis. Line B is as follows: Ekman's school, numerical modeling of a barotropic and then baroclinic ocean, modeling of individual basins with high resolution, and modeling of the World Ocean with high resolution and allowance for ice cover. This paper briefly reviews and analyzes the results of each of the abovementioned stages. The author sympathizes with line A because he believes that the most realistic results for characteristics of scien tific interest are obtained by a synthesis of models and measurement data, with an optimal choice of the model integration time. Unfortunately, studies in both directions have often used unreasonably long integra tion times (50-100 years). In this case, models even with a high resolution (0.1° of grid spacing) become inad equate and analyze strange characteristics, such as meridional overturning or kinetic energy averaged over the entire column of the world ocean, which actually have no scientific significance.