Several classes of active and selective catalysts for butene skeletal isomerization are already known. The most promising ones are those, based on zeolites or zeotypes. These materials suppress side reactions, and they are stable on stream. This paper shows why this is so : dimerization and oligomerization (the latter being the source of deactivation) are suppressed when a proper structure of the microporous catalyst is chosen; i.e., noncrossing, 10-membered ring (10-MR) channels. It is concluded, at variance with most of the early literature that the prevailing mechanism, which has to be induced by the catalyst, is a monomolecular mechanism. The bimolecular mechanism should be suppressed as much as possible.