The percentage of equivalence classes generated for the component, which have been exercised by a test case suite.
BS 7925-1.British Computer Society Specialist Interest Group in Software Testing (BCS SIGIST)

Another metric that requires the test organisation to have proper planning and records keeping procedures. The reason for planning is that to calculate the percentage, they need to know the total number of equivalence partitions or class they will need to test.

The ideal is to have the percentage go up every time testing occurs. However if the development has not been estimated properly, the number of equivalence classes is likely to keep changing, thus distorting the metric.

Standards have to be enforced so that everyone uses the same definition of equivalence partition or class. Additionally that there are agreed test criteria completion details.

Thus more accurate estimation and agreed standards make for a more accurate metric.

A similar metric is boundary value coverage.