We present Harvey, a tool for random testing in answer-set programming (ASP) that allows to incorporate constraints to guide the generation of test inputs. Due to the declarative nature of ASP, it can be argued that there is less need for testing than in conventional software development. However, it is shown in practice that testing is still needed when more sophisticated methods are not viable. Random testing is recognised as a simple yet effective method in this regard. The approach described in this paper allows for random testing of answerset programs in which both test-input generation and determining test verdicts is facilitated using ASP itself: The test-input space is defined using ASP rules and uniformity of test-input selection is achieved by using XOR sampling. This allows to go beyond simple random testing by adding further ASP constraints in the process.