The number of defects identified in a component or system divided by the size of the component or system (expressed in standard measurement terms, e.g. lines-of code, number of classes or function points).