Important parameters when negotiating a ship and also when commissioning it are the speed attained at a specific engine power having a specific (usually nominal) RPM of the engine. From the very beginning of implementation of ship model tests, efforts were made to establish a link between the results of model tests and full-scale measurements. By testing an increasing number of ship models and having the increased number of full-scale tests, and by comparing the predicted speeds obtained by model tests and the speeds achieved by full-scale ships, preconditions have been created for determining the correlation for actual ships and thereby improve the predictions for new ships. It is often the case that the success of a new project and meeting of contractual obligations depend on a well-predicted speed. The importance of continuous research of the correlation by investigating the correlation methods used when correlating the results obtained by ship model tests and the results of propulsion characteristics obtained by full-scale tests is pointed out. The most important methods applied for the transfer of model results to full-scale ship values are presented. Their advantages and disadvantages are discussed. Experiments with ship models that are aimed at forecasting the speed, the required power and the engine speed are briefly described. Comments are given concerning the accuracy of the measured quantities. Full-scale propulsion tests are also briefly described, and the accuracy of the measurement of particular quantities is considered. The method used for defining the correlation lines for the correction of the predicted power and engine speed is presented. A correlation line is made on the basis of the available ship model data obtained in model tests performed in the Brodarski Institute’s testing facilities and the related full-scale tests (for measuring miles). Within this framework the possibilities for better defining of correlation (correlation coefficients) have been studied. Finally, comparison between the results obtained and the results presented were made 20 years ago in Brodarski Institute and the comments and suggestions for further work are given.