Two studies examined abstract conditional reasoning. First, participants at three grade levels (grade 8, college, university) were informed that a problem corresponding to Affirmation of the consequent (P implies Q, Q is true) had no certain response and were asked to justify this conclusion, for both concrete and abstract premises. Results showed an increase in use of formal justifications with grade. The second study looked at the effects of embedding abstract premises into realistic or fantasy contexts with participants at grades 8 and 11. Results show that abstract reasoning is facilitated by realistic context. These results support the idea that such reasoning may represent a qualitative change in reasoning abilities and that its development relies on appropriate access to empirical knowledge despite the abstract nature of premises.