ABSTRACT: The world appears causal in the sense that the result of a measurement may depend on the past history of the observed system, but not on what the experimenter will do with the system after the measurement. This raises the question whether noncausality at a macroscopic level would necessarily lead to an "unreasonable" world. The study of a model world with axiomatically well-specified properties shows that noncausal systems can be discussed in a logically consistent manner so that noncausality might well exist in the real world as a weak, but so far overlooked, effect.