We present Z-log – a practical system that employs the system-Z [13] semantics. Z-log incurs polynomial cost for compilation and entailment in the horn and q-horn [2] cases. Z-log’s complexity is intractable in the unrestricted case – but intractable in the number of defaults that cause the violation of the q-horn property. We present here initial performance results over two alternative rulesbases. The results indicate that Z-log currently scales to problems on the order of 1000’s of propositional rules when the rules are in q-Horn form. We shall be applying Z-log in cognitive disease diagnosis.