Fibonacci’s treatment of so-called recreational problems in his Liber Abbaci has been interpreted as an early episode both in the history of systems of linear equations, and in the history of negative numbers. However, these problems are also interesting in their own right. We discuss some of the conventions which seem to have governed these problems. By considering certain pairs of problems, where one problem is unsolvable and its partner is solvable, we show that Fibonacci went to a significant effort (...) to conform to these conventions. We also examine the methods which he could have used to construct his problems. (shrink)