This study aims to gain greater insight into scheduling expertise by comparing the work of experts and novices when designing a university timetable. We assumed that the scheduling activity would take place within two dual spaces: the constraints space (CS) and the objects space (OS). Constraints are defined in the strictest sense as relations between variables that cannot be represented in the solution (timetable), whereas objects are constraint satisfactions that can be thus represented. The study shows that experts were more likely than novices to use external representations as activity support. They satisfied many constraints with partially defined objects. On the contrary, novices devoted a long time to managing constraints in their heads before defining only fully specified objects (concrete objects). The objects space could be a suitable activity support for experts. Novices, on the other hand, could benefit from support in managing constraints and translating constraints into objects. © 2012 Wiley Periodicals, Inc.