The formulation of quantum mechanics on topologically nontrivial spaces is discussed. It is pointed out that the "obstacles" are represented by cohomology groups and not only by pi-1 (M) as usually stated. Some widespread errors and misunderstandings are cleared up.