In connection with the development of submarine project A14 and A17, parametric study programs were used during initial design, i.e. study-, conceptual and preliminary design phases. This was done to establish the future submarines size, as well as to determine the composition of submarine parameters in order to obtain the best performance available for a given cost. The differences between the final size of the submarine projects and the early parameter study results were however large. As a result the parametric approach was not used or further developed after the submarine project A17. Instead the traditional type ship development and design method was again used for submarine type A19 Gotland, i.e. an evolutionary development of previous submarine, and in this case the submarine type A17 Västergötland was used. In the present study the causes of the differences has been analysed and the result indicates that there were systematic errors in the application of the parametric design programs TC 112A and TC 117A/E. Shortcomings can be attributed to underestimates of weights for the hull, including the pressure hull, hull equipment and ballast weight to get in balance. In addition, underestimates of free volumes in gangways and spaces and volumes for shock distance, service, maintenance and general volumes for systems, equipment and components such as cabling, ventilation, piping etc. There were also some softer reasons for the smaller size of the parameter sets, such as an ambition to design smaller submarine with the same performance and system effect in the hope of a less costly submarine. Never the less, in summary, this gave sizing mistakes both in respect of volumes and weights and therefore the generated submarines became 6-22% too small compared with the submarine type A17 Västergötland as built. Generally, the consequences of these systematic mistakes, unless corrected early enough, are either high costs for redesign or less effective performance in operation and maintenance.