Abstract The article demonstrates that the rules and practices of cabinet formation and investitures should be taken into account to better grasp the variety of semi-presidentialism in Europe. This is extremely important, since semi-presidentialism as a constitutional system of government (primarily according to a minimalist approach to the definition) is the most common form of inter-institutional and political relations in European countries. The former is most often understood as a constitutional design of inter-institutional relations with a president popularly elected for a fixed term, as well as with a cabinet headed by a prime minister who are collectively responsible to parliament. Thus, not only presidents and parliaments, but cabinets too play a crucial role in the constitutional and political practice of semi-presidentialism, since the latter are collectively responsible to parliaments (or simultaneously to presidents), but are characterised by distinctive parameters of formation. The assumption and hypothesis are that options of cabinet formation and inter-institutional relations in this regard can structure European semi-presidentialism, even without affecting the definition of this constitutional design, but probably depending on the roles and powers of presidents and parliaments in cabinet formation, as well as types of semi-presidentialism regarding the consideration of who can dismiss the cabinet. Based on comparison and systematisation of the cases of European semi-presidentialism, it is justified that cabinet formation (including within various types and consequences of parliamentary votes of investiture in cabinets, as well as their absence) is typically focused on a junction of relations between presidents and parliaments, and are likely to serve as a classification indicator of semi-presidentialism. This is important for constitutional engineering, since detailing the optionality of semi-presidentialism as a constitutional design and system of government, particularly regarding cabinet formation, should extend the horizons, as well as systematise the idea of the options and effects of various institutional designs (in addition to presidentialism and parliamentarism) and political regimes (including democratic, autocratic and hybrid).