There have been several suggestions on how Rational Choice Theories (RCT) for educational decisions could be tested directly. In this empirical analysis, it is argued that these testing strategies have shortcomings, and therefore a different testing strategy is suggested. An innovative specification of a broad version of an RCT for educational decisions is applied, directly taking the utility-maximising behaviour of the actors into account. By using this improved testing strategy, it is demonstrated for young people residing in German-speaking Switzerland that rational decisions play an important role in educational decisions. They calculate the subjective expected benefit of every educational option at the end of compulsory school that, in their view, is relevant to the continuation of their school and vocational training. Panel data show that they ordinally rank the educational options considered according to their benefit and that they choose the option that promises the greatest benefit. For young people in different school branches, it can be seen that the purposive-rational evaluation of the educational options considered and the selection of the educational option with the highest subjectively expected utility are essential mechanisms of educational decision-making characterised by procedural rationality.