Simple pacing has often been suggested as an explanation of the 'megalithic yard'. While such a hypothesis has the virtue of simplicity, evidence is presented here showing it to be inconsistent with actual stone circle data. By means of simulations, the pacing hypothesis is tested against the stone diameters given by Thom (9167), and may be rejected at a significance level of about 3per thousand. Further simulations suggest that any satisfactory explanation of the unit should not involve variations in its value between sites greater than about 2per thousand or 3per thousand (standard deviation).