The problem of parsing ambiguous structures concerns (i) their representation and (ii) the specification of mechanisms allowing to delay and control their evaluation. We first propose to use a particular kind of disjunctions called controlled disjunctions: these formulae allows the representation and the implementation of specific constraints that can occur between ambiguous values. But an efficient control of ambiguous structures also has to take into account lexical as well as syntactic information concerning this object. We then propose the use of unapt quasi-trees specifying constraints at these different levels. The two devices allow an eff