Two procedures are presented: one for diagnosis of single-gate failures and the other for deriving equivalence classes of single faults. Both procedures require use of a diagnostic table which lists gate sensitivities and primary output values for all input patterns in a set of tests which contains at most U tests, where U is an upper bound derived in the text. A gate is defined to be an n-input single-output combinational circuit (n ≥ 1) and a failure is any detectable transformation of the correct gate function. The information contained in a diagnostic table can be used to locate a faulty gate and describe its failed function. Equivalence classes are derived from a diagnostic table by first specifying the primary output values in accordance with a theorem given in the text and then deriving the corresponding gate functions. Examples are given to illustrate both the procedures.